EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH (CERN) CERN-LHCb-DP-2014-001 September 11, 2014 Performance of the LHCb Vertex Locator LHCb VELO Group † Abstract The Vertex Locator (VELO) is a silicon microstrip detector that surrounds the proton-proton interaction region in the LHCb experiment. The performance of the detector during the first years of its physics operation is reviewed. The system is operated in vacuum, uses a bi-phase CO 2 cooling system, and the sensors are moved to 7 mm from the LHC beam for physics data taking. The performance and stability of these characteristic features of the detector are described, and details of the material budget are given. The calibration of the timing and the data processing algorithms that are implemented in FPGAs are described. The system performance is fully characterised. The sensors have a signal to noise ratio of approximately 20 and a best hit resolution of 4 μm is achieved at the optimal track angle. The typical detector occupancy for minimum bias events in standard operating conditions in 2011 is around 0.5%, and the detector has less than 1% of faulty strips. The proximity of the detector to the beam means that the inner regions of the n + -on-n sensors have undergone space-charge sign inversion due to radiation damage. The VELO performance parameters that drive the experiment’s physics sensitivity are also given. The track finding efficiency of the VELO is typically above 98% and the modules have been aligned to a precision of 1 μm for translations in the plane transverse to the beam. A primary vertex resolution of 13 μm in the transverse plane and 71 μm along the beam axis is achieved for vertices with 25 tracks. An impact parameter resolution of less than 35 μm is achieved for particles with transverse momentum greater than 1 GeV /c. To be submitted to Journal of Instrumentation † Authors are listed on the following page. arXiv:1405.7808v2 [physics.ins-det] 10 Sep 2014
Transcript
EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH (CERN)
CERN-LHCb-DP-2014-001September 11, 2014
Performance of the LHCb VertexLocator
LHCb VELO Group†
Abstract
The Vertex Locator (VELO) is a silicon microstrip detector that surrounds theproton-proton interaction region in the LHCb experiment. The performance of thedetector during the first years of its physics operation is reviewed. The systemis operated in vacuum, uses a bi-phase CO2 cooling system, and the sensors aremoved to 7 mm from the LHC beam for physics data taking. The performance andstability of these characteristic features of the detector are described, and details ofthe material budget are given. The calibration of the timing and the data processingalgorithms that are implemented in FPGAs are described. The system performanceis fully characterised. The sensors have a signal to noise ratio of approximately 20and a best hit resolution of 4µm is achieved at the optimal track angle. The typicaldetector occupancy for minimum bias events in standard operating conditions in 2011is around 0.5%, and the detector has less than 1% of faulty strips. The proximityof the detector to the beam means that the inner regions of the n+-on-n sensorshave undergone space-charge sign inversion due to radiation damage. The VELOperformance parameters that drive the experiment’s physics sensitivity are also given.The track finding efficiency of the VELO is typically above 98% and the moduleshave been aligned to a precision of 1µm for translations in the plane transverse tothe beam. A primary vertex resolution of 13µm in the transverse plane and 71µmalong the beam axis is achieved for vertices with 25 tracks. An impact parameterresolution of less than 35µm is achieved for particles with transverse momentumgreater than 1 GeV/c.
To be submitted to Journal of Instrumentation
†Authors are listed on the following page.
arX
iv:1
405.
7808
v2 [
phys
ics.
ins-
det]
10
Sep
2014
LHCb VELO group
R. Aaij1, A. Affolder2, K. Akiba3, M. Alexander4, S. Ali1, R.B. Appleby5, M. Artuso6,A. Bates4, A. Bay7, O. Behrendt8, J. Benton9, M. van Beuzekom1, P.M. Bjørnstad5,G. Bogdanova10, S. Borghi5, A. Borgia6, T.J.V. Bowcock2, J. van den Brand1, H. Brown2,J. Buytaert8, O. Callot11, J. Carroll2, G. Casse2, P. Collins8, S. De Capua5, M. Doets1,S. Donleavy2, D. Dossett12, R. Dumps8, D. Eckstein13, L. Eklund4, C. Farinelli1, S. Farry2,M. Ferro-Luzzi8, R. Frei7, J. Garofoli6, M. Gersabeck5, T. Gershon12, A. Gong14,H. Gong14, H. Gordon8, G. Haefeli7, J. Harrison5, V. Heijne1, K. Hennessy2,W. Hulsbergen1, T. Huse15, D. Hutchcroft2, A. Jaeger16, P. Jalocha17, E. Jans1, M. John17,J. Keaveney18, T. Ketel1, M. Korolev10, M. Kraan1, T. Lastovicka19, G. Lafferty5,T. Latham12, G. Lefeuvre6, A. Leflat10, M. Liles2, A. van Lysebetten1, G. MacGregor5,F. Marinho20, R. McNulty21, M. Merkin10, D. Moran22, R. Mountain6, I. Mous1,J. Mylroie-Smith2, M. Needham23, N. Nikitin10, A. Noor2, A. Oblakowska-Mucha24,A. Papadelis1, M. Pappagallo4, C. Parkes5, G.D. Patel2, B. Rakotomiaramanana7,S. Redford8, M. Reid12, K. Rinnert2, E. Rodrigues5, A.F. Saavedra25, M. Schiller1,O. Schneider7, T. Shears2, R. Silva Coutinho12, N.A. Smith2, T. Szumlak24, C. Thomas17,J. van Tilburg1, M. Tobin7, J. Velthuis9, B. Verlaat1, S. Viret26, V. Volkov10, C. Wallace12,J. Wang6, A. Webber5, M. Whitehead12, E. Zverev10.
1Nikhef National Institute for Subatomic Physics, Amsterdam, Netherlands2Oliver Lodge Laboratory, University of Liverpool, Liverpool, United Kingdom3Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil4School of Physics and Astronomy, University of Glasgow, Glasgow, United Kingdom5School of Physics and Astronomy, University of Manchester, Manchester, United Kingdom6Syracuse University, Syracuse, NY, United States7Ecole Polytechnique Federale de Lausanne (EPFL), Lausanne, Switzerland8European Organization for Nuclear Research (CERN), Geneva, Switzerland9H.H. Wills Physics Laboratory, University of Bristol, Bristol, United Kingdom10Institute of Nuclear Physics, Moscow State University (SINP MSU), Moscow, Russia11LAL, Universite Paris-Sud, CNRS/IN2P3, Orsay, France12Department of Physics, University of Warwick, Coventry, United Kingdom13Deutsches Elektronen-Synchrotron, Hamburg, Germany14Center for High Energy Physics, Tsinghua University, Beijing, China15Department of Physics, University of Oslo, Oslo, Norway16Physikalisches Institut, Ruprecht-Karls-Universitat Heidelberg, Heidelberg, Germany17Department of Physics, University of Oxford, Oxford, United Kingdom18Vrije Universiteit Brussel, Brussel, Belgium19Institute of Physics, Academy of Sciences of the Czech Republic, Prague, Czech Republic20Universidade Federal do Rio de Janeiro, Campus UFRJ - Macae, Rio de Janeiro, Brazil21School of Physics, University College Dublin, Dublin, Ireland22Universidad Autonoma de Madrid, Spain23School of Physics and Astronomy, University of Edinburgh, Edinburgh, United Kingdom24AGH - University of Science and Technology, Faculty of Physics and Applied Computer Science,Krakow, Poland25School of Physics, University of Sydney, Sydney, Australia26Universite de Lyon, Universite Claude Bernard Lyon 1, CNRS-IN2P3, Institut de Physique Nucleairede Lyon, Villeurbanne, France
LHCb [1] is an experiment dedicated to heavy flavour physics at the LHC. Its primaryaim is to discover new physics through precision studies of CP violation and rare decaysof beauty and charm hadrons. The detector is comprised of the Vertex Locator (VELO),silicon strip and straw chamber trackers, a warm dipole magnet, ring imaging Cherenkovparticle identification systems, calorimeters and a muon detection system.
The LHCb Vertex Locator [2] is a silicon microstrip detector positioned around theproton-proton interaction region. The VELO provides measurements of track coordinateswhich are used to identify the primary interaction vertices and the secondary vertices thatare a distinctive feature of b- and c-hadron decays. The VELO was designed to optimisethe LHCb physics programme in the following ways:
• Angular coverage. The VELO is designed to cover the forward region, such that alltracks inside the nominal LHCb acceptance of 15–300 mrad cross at least three VELOstations. In this way the detector fully reconstructs roughly 27% of bb productionfor 7 TeV proton-proton centre-of-mass collisions, while covering just 1.8% of thesolid angle [3, 4]. The VELO also reconstructs tracks in the forward direction andbackward directions which do not have momentum information, but are useful toimprove the primary vertex reconstruction.
• Triggering. The reconstruction of the primary vertex and the displaced secondarydecay vertex of a heavy flavour hadron in the VELO is a key ingredient of the highlevel trigger which reduces the event rate from a 1 MHz event rate to a few kHz [5].
• Efficient reconstruction. LHCb has studied decay modes with up to six chargedtracks in the final state [6]. This relies on the highly efficient cluster reconstructionin the VELO since track reconstruction efficiency losses are transmitted as the sixthpower. The cluster reconstruction efficiency in the VELO is paramount, both forthe selection of those tracks, as six measurements per track are required, and forefficient pattern recognition and fake track rejection.
• Displaced tracks and vertices. Excellent vertex resolution is essential to theLHCb physics programme. Most analyses rely heavily on selection cuts on thedistance with which tracks approach the primary vertex (impact parameter) and onthe displaced vertex reconstruction with the VELO to identify the signal channels.The impact parameter resolution was optimised by positioning the VELO sensorsas close to the LHC beam as permitted by safety consideration, having a smallinter-strip pitch at the inside of the sensors, and minimising the amount of materialtraversed by a particle before the first measured hits in the VELO.
• Decay time. The decay time of a particle is obtained from the measurement ofits flight distance in the VELO. This is required for lifetime measurements and,critically, for time-dependent measurements in the rapidly oscillating B0
s–B0s meson
system [7,8].
1
Figure 1: (top left) The LHCb VELO vacuum tank. The cut-away view allows the VELO
sensors, hybrids and module support on the left-hand side to be seen. (top right) A photograph of
one side of the VELO during assembly showing the silicon sensors and readout hybrids. (bottom)
Cross-section in the xz plane at y = 0 of the sensors and a view of the sensors in the xy plane.
The detector is shown in its closed position. R (Φ) sensors are shown with solid blue (dashed
red) lines. The modules at positive (negative) x are known as the left or A-side (right or C-side).
The VELO contains a series of silicon modules arranged along the beam direction,see Fig. 1. A right-handed co-ordinate system is defined with z along the beam-axis intothe detector, y vertical and x horizontal. Cylindrical polar co-ordinates (r, θ, φ) are alsoused. The region of the detector at positive (negative) z values is known as the forward(backward) or downstream (upstream) end.
The sensors are positioned only 7 mm from the LHC beams. This is smaller than theaperture required by the LHC beam during injection. Hence, the detector is produced intwo retractable halves. There is a small overlap between the two detector halves whenclosed. This aids alignment and ensures that full angular coverage is maintained. Theposition of the VELO halves are moveable in x and y and the VELO is closed at thebeginning of each fill such that it is centred on the interaction region.
Approximately semi-circular silicon sensors are used. Each module contains one r andone φ coordinate measuring sensor, known as R and Φ sensors and shown schematically inFig. 2. The inter-strip pitch varies from approximately 40 to 100µm across the sensor. Thestrips are read out from around the circumference of the sensor through the use of routinglines on the sensor. The sensors are read out using the Beetle [9] analogue front-end ASIC,operated with a 40 MHz input event sampling rate. The signals are digitised and processed
2
Figure 2: Schematic representation of an R and a Φ sensor. The R sensor strips are arranged
into four approximately 45◦ segments and have routing lines perpendicular to the strips. The
Φ sensor has two zones with inner and outer strips. The routing lines of the inner strips are
orientated parallel to the outer strips.
to form clusters in an FPGA-based readout board known as the TELL1 [10], before beingpassed to the high level trigger. More details on the readout chain are given in Sect. 3.1
There are 21 standard modules in each VELO half. Two further modules, known asthe pile-up system, containing R sensors only are located in the most upstream positions.
Owing to the proximity of the detector to the beam, the VELO is exposed to a highradiation fluence and radiation tolerant oxygenated n+-on-n sensors, consisting of ann-type implant on an n-type bulk with a backplane p+-type implant, are employed. Oneof the most upstream modules uses n+-on-p silicon (one R and one Φ sensor pair). Theseare the only n+-on-p sensors in operation at the LHC and were installed as this technologyis a leading candidate for use in the LHC upgrades. All sensors were fabricated by MicronSemiconductor.1
The detectors are mounted in a vacuum vessel and are located in a secondary vacuumseparated from the LHC machine vacuum by an RF-box. The surfaces facing the beam are0.3 mm thick corrugated sheets, known as the RF foil. These RF-boxes and foil providethree functions: they provide shielding against RF pickup from the LHC beams, guidewakefields to prevent impedance disruptions to the LHC beams, and protect the LHCvacuum from outgassing of the detector modules. The detector is located upstream of theLHCb dipole magnet in a region with a negligible magnetic field.
1Micron Semiconductor Ltd., 1 Royal Buildings, Marlborough Road, Lancing Business Park, Lancing,Sussex, BN15 8SJ, UK.
3
This paper reports on the performance of the VELO detector over the first period ofLHC physics operation. The first proton-proton collisions occurred in November 2009 atbeam energies of 450 GeV, with the first 3.5 TeV beam collisions occurring in March 2010.The beam energy was raised further to 4 TeV in April 2012. All physics data were recordedwith a 50 ns minimum bunch spacing. LHCb recorded integrated luminosities of 0.04 fb−1
in 2010, 1.11 fb−1 in 2011 and 2.08 fb−1 in 2012. The first period of LHC operations endedin February 2013, when the LHC entered a shutdown for an upgrade to increase the beamenergy. The LHCb VELO performance results in this paper are primarily given on 2011data, the first year in which the instantaneous luminosity reached, and exceeded, thenominal design value. Section 2 describes the performance of the component subsystems inthe VELO. The calibration of the timing, gain and processing parameters is described inSect. 3, along with the detector monitoring strategy and the simulation and reconstructionsoftware. Sections 4 and 5 provide the system performance results, with the former sectionproviding detector performance related results and the latter results on quantities moredirectly related to physics performance.
4
2 Subsystem performance
The performance of individual subsystems from the VELO are described in this section.The section starts with a description of the commissioning stages of the detector andthen describes the performance of the vacuum, cooling, low and high voltage and motionsystems. The section ends with information on the material budget of the VELO.
2.1 Commissioning results
The subsystems and overall operations of the VELO were extensively commissioned beforethe first LHC beam collisions. After production, the commissioning of the VELO consistedof four main stages: testing during assembly, operation in a test beam, commissioningpost-installation at the LHCb experimental area, and testing with beam-absorber collisions.
The detectors underwent extensive quality assurance tests at the production sites andafter delivery for system assembly, utilising components of the final readout system [11].During detector assembly (Fig. 1 shows a photograph of the system at this stage) eachmodule was operated with a complete readout slice, simulating as closely as possiblethe final vacuum, cooling, and powering conditions of the VELO. The main mechanicaldifference from the final system was the absence of the RF foil. This first testing stepallowed a complete “fingerprint” of the pedestal and noise map of each module to beproduced, along with the commissioning of the zero suppression algorithms running in theTELL1 DAQ boards (see Sect. 3.3).
In the second step, a partially assembled VELO half was taken to the 120 GeVpion/muon beam at the North area SPS test beam facility at CERN. Complete track andvertex reconstruction was performed, using the products of interactions of the beam in leadtargets placed in similar positions to that of the LHC beam interaction region [12]. Thisallowed the data acquisition and software chain to be debugged, in addition to performinga full speed test of significant components of the LHCb readout chain. It also served as a“dress rehearsal” for the transportation of the VELO using a truck with special tractionand speed control. The VELO half was mounted on four shock absorbers on a trailer withspecial suspension, and the transport motion was fully logged using accelerometers.
The third stage of the commissioning was carried out after the installation of the VELOat the LHCb experimental area. Due to the proximity of the silicon sensors (≤ 1 mm)to the 1 m long corrugated RF foil, and the insertion procedure of the detector halvesinto the RF foil relying purely on the manufacturing and assembly tolerances, it wasnecessary to employ special monitoring procedures. Each module slot was tested with anoversized dummy sensor model for contact with the foil, and a test insertion was performedwith a genuine module. Characteristic IV curves of all sensors were taken at severaloccasions during the transport, movement and installation procedures, and it was verifiedthat no damage had occurred. Following this, the cabling of the detector was verified byapplying custom test-pulse patterns to each of the 5632 readout links, and verifying thecorrect pattern was read back. Each link was tuned to select the optimal ADC samplingpoint using test-pulses (see Sect. 3.2.1). The digitisation uniformity was confirmed by
5
injecting sine wave pulses with an amplitude close to the full dynamic range and a periodmatching the pulse train from the front-end ASICs. In addition, the noise and pedestalswere measured, and compared with the “fingerprint” recorded during the assembly andcommissioning process.
In the fourth and final commissioning step, a special method was found to studygenuine tracks in the detector before the start of LHC collisions. The synchronisation testsof the LHC beam were utilised in which, during the initial phase of each test, the LHCproton beam was collided with a beam absorber at the end of the transfer line between theCERN SPS and the LHC about 340 m from the LHCb cavern. This allowed the VELOto reconstruct the first tracks from the LHC machine [13]. The particles produced bythe proton interactions in the absorber, and by their re-interaction, were detected by theLHCb experiment and were used to commission the detector. These injection tests wereperformed in 2008 before the first circulating beam in the LHC and in 2009 before the firstproton-proton collisions. These data were used to set up the timing and the alignment ofthe system, and for commissioning of the control, reconstruction and monitoring software.They were of particular importance to the VELO as it is not possible to collect sufficientcosmic ray data for commissioning due to the horizontal geometry, and hence provided thefirst full system test of the VELO. With this method about 50k tracks were reconstructedin LHCb. This was sufficient to determine that the modules were displaced by less than10µm from their surveyed positions (see Sect. 5.2). The alignment of the modules wasdetermined with 5µm precision for x and y translation and 200µrad for the rotationsaround the z-axis [14]. In addition, the tracks traversing from one VELO half to theother could be used, by evaluating the mismatch of the two segments, to measure thedistance between the halves to a precision of 100µm. The time alignment of the VELOwas determined to a precision of 2 ns, and measurements made of the signal to noise andcluster finding efficiency. These commissioning stages allowed LHCb to be ready for animmediately successful start to physics data taking after the first LHC beams collided inNovember 2009.
2.2 Vacuum stability
The VELO vacuum vessel is composed of two sections: the first is part of the beam volumeof the LHC, and the other contains the detector modules. The two sections are separatedby a vacuum tight RF-box, welded from a 0.3 mm thick corrugated AlMg3 foil with 0.5 mmthick side walls which encapsulates each detector half. The differential pressure betweenthe beam and detector volume should always remain below 5 mbar to protect the foilagainst irreversible deformation. Consequently, the pump-down and gas venting proceduresare very delicate operations that are controlled by a programmable logic controller (PLC),based on the readings from three membrane switches that have preset trigger values.The PLC continuously monitors the performance of the system and takes, when needed,appropriate actions. A PVSS2 project monitors and archives all relevant system variables
2PVSS is a Supervisory Control and Data Acquisition software package developed by ETM professionalcontrol GmbH.
6
Time (hrs)10 20 30
Pres
sure
(m
bar)
-910
-810
-710PE 412
PE 422
Figure 3: Illustration of the influence of the proton beams on the beam volume pressure. The
readings of the beam volume (PE 412) and detector volume (PE 422) are shown, and the increase
in pressure in the beam volume coinciding with the beam injections (each containing 2× 1014
protons) at t = 6 hours can be seen. The beams were dumped 24 hours later.
in a database, that is common to the whole LHCb experiment [15]. For redundancypurposes, two sets of roughing, turbo-molecular and ion pumps are implemented in thesystem. The pressure in the detector volume is around 2 × 10−7 mbar, while the beamvolume is at 1× 10−9 mbar in the absence of circulating beams. Under the influence of1380 bunches per beam this pressure can rise to 5×10−9 mbar. The influence of the protonbeams on the pressure of the beam volume can clearly be seen in Fig. 3.
2.3 Cooling performance
The VELO modules are cooled by means of an evaporative system using CO2 as coolant.The cooling plant, which is located outside the radiation zone, has two chillers filled withhydrofluorocarbon (R507a): a 2.5 kW water-cooled main chiller and a 1 kW air-cooledbackup chiller, operating at −40 and −25◦C, respectively (see Ref. [16]). The actualcooling system of the detector consists of two parallel, largely independent, loops: one foreach detector half. The CO2 is subcooled by the main chiller and subsequently pumped tothe corresponding detector via a 55 m long transfer line. The operation principle is thatof a bi-phase accumulator controlled loop, in which the evaporative temperature in thedetector is regulated by controlling the pressure of the saturated mixture of vapour andliquid in the accumulator. The cooling system is controlled by a PLC. In total 31 pressureand 192 temperature sensors are installed throughout the system. The PLC controls thesystem via 24 proportional, integrating and differentiating loops based on a subset of thesereadings. Alarm handling tasks are executed continuously and if appropriate the systemswitches to the backup chiller or a spare pump. A PVSS project monitors and archives
7
all relevant system variables in an Oracle database that is common to the whole LHCbexperiment. A separate heating system takes care of maintaining the temperature of themodule base at 20◦C to minimise deformations.
The default evaporator temperature is −28◦C. Due to thermal gradients between thecooling blocks and the components of the hybrid the silicon temperatures are (−7± 2)◦C.From the moment the front-end electronics is switched on, it takes 3 minutes till thehybrid temperatures have stabilised. The hybrid temperatures have been shown to bestable within 0.1◦C over a period of four weeks.
The operation of the cooling system has been very smooth. The only major interventionrequired was the replacement of the insulation of the transfer lines in 2011 to eliminatethe formation of ice. During operations in both 2009 and 2011 an increase of ∼1◦C wasobserved in one detector half and found to be due to clogging of filters: the filters werereplaced during the following winter shutdowns.
During injection and ramp of the beams, the temperatures of the RF-boxes increase by0.5◦C. Subsequently the temperatures of the RF-boxes decrease by 1.5◦C when the VELOis closed.
2.4 Low voltage and high voltage
Each silicon sensor has its own hybrid with separate low voltage (LV) and high voltage(HV) supplies. In addition, repeater boards are located directly outside the VELO vacuumtank, which have their own positive and negative supplies. Both the LV and HV powersupplies are inside the counting-house shielded from the radiation zone of the detector.
The LV system is based on the A3009 EASY low voltage module, with compatiblecrates and mainframe controller, manufactured by CAEN.3 The main system performanceissue has been with the LV Anderson connectors used in the system. Due to oxidationa voltage drop across the connector can occur. This requires careful monitoring as thedissipated power caused severe thermal damage on one of the connectors during operation.The connectors were replaced by CAEN during the shutdown at the end of 2011.
The HV system is based on the EHQ F607n-F 16-channel HV module from Iseg.4
Careful monitoring of the HV parameters is a key requirement as the current and depletionvoltage change with radiation damage (see Sect. 4.6). The main performance issue waswith the control of the HV supplies. The supplies occasionally get stuck when ramping involtage, requiring the software commands to be resent or control software restarted.
Voltage and current monitoring are implemented for both LV and HV systems andthey are connected to an FPGA based interlock safety system. All HV and LV modulescan have their channels switched off quickly (< 1 s) by the interlock system in case ofproblems. The primary uses of the interlock system have been in the case of power cutsor disruptions, and interruptions to the chilled water supply that is used by the coolingsystem.
Figure 4: The horizontal position of LHC collision vertices as reconstructed by the A-Side (left)
and C-Side (right). These online screenshots of data illustrate the “stop-measure-move” cycle of
the closing procedure, where each peak in the distribution corresponds to a “stop”, as well as
the degradation of vertex resolution with opening distance.
2.5 Motion performance
The first active silicon strips are brought to within 8.2 mm of the LHC beams, with theinner surface of the RF foil at a 5.5 mm radius. To ensure detector safety during beaminjection and adjustments, each VELO half is retracted ∼ 29 mm in the horizontal planeand is only closed once stable-beam conditions are declared. The VELO halves are movedusing radiation hard stepper motors, and the motion towards the final position is alwaysdone from the same direction in order to minimise the effects of mechanical backlash. Theposition is read from resolvers mounted on the motor axes. The reproducibility of theposition has been measured to be better than 10µm. The motion system is controlled by aPLC which performs many safety checks to prevent unwanted movements due to hardwarefailures or corruption of the destination position calculated by the closing procedure.
An automated closure procedure has been developed to position the VELO halvesaround the beams, whilst taking into account the current beam conditions. It usesinformation about the background level, the response of the hardware and the positions ofthe beams to make informed decisions. During closure, the LHCb trigger accepts 500 Hzof events which are required to contain at least one track in the VELO. This rate ensuresthat a new beam-position measurement, consisting of 400 reconstructed vertices, can beacquired in 1–2 s, even in the most open position. Figure 4 shows an example of thedistribution of vertices during a typical closure.
Upon receiving notification of stable beams, the two VELO halves are moved in a seriesof “stop-measure-move” steps from open to closed. A vertical adjustment is also madeat each of the steps, if the movement required to centre the VELO around the beams isgreater than 20µm. Three criteria are checked at all times during and after closure:
1. the total silicon bias current from the 44 sensors in each half of the VELO is lessthan 1000µA above the dark current. This limit corresponds to 10% occupancy ofionising particles in all sensors for several seconds;
9
2. the Beam Conditions Monitors [17] — each consisting of eight CVD-diamond radia-tion monitors placed adjacent to the beam pipe — are all functional and reportingfluences less than 5% of the threshold which triggers an LHC beam dump;
3. the reconstructed beam position in x and y, as measured by the two halves, agreegiven the known opening distance at any given step. The observed beam width mustbe acceptable given the opening distance and the expected resolution thereof.
By considering the two independent beam profiles compiled by each half, the VELO isobserved to close symmetrically to an accuracy of better than 4µm.
Once closed the monitoring continues throughout the data taking. Apart from requiringthe above conditions, a slow drift of the beams is protected against by prompting theVELO to open if the beams move by more than 300µm. The LHC beam orbit has shownexcellent stability. Consequently, the VELO is permitted to remain closed for up to 20minutes if the reconstruction of the beam profiles is interrupted. This happens, for example,if the LHCb data acquisition is paused (e.g. for a run change or reconfiguration). Duringthis grace period, movement of the beam is monitored by the LHC beam position monitorswhich are horizontal and vertical wire pickups located at z = ±22 m of the collision region.A deviation of 200µm in any of these readings will trigger the VELO to open.
The initial use of the system was performed with careful manual checking and controlof the closings. The automated closing procedure was then put in place. From thedeclaration of stable beams the VELO takes, on average, 210 s to close. Of this 160 secondsis due to the motion from the open to closed positions. During the operations in 2010–2012 approximately 750 closing procedures were performed and 0.9% of the stable-beamintegrated luminosity delivered by the LHC was lost due to this detector-safety procedure.This number is in line with the expectation. Only minor performance problems have beenencountered and these have been addressed with changes to the closing procedure, safetychecks and motion PLC code.
2.6 Material description
The minimisation of the material budget of the detector is important for the physicsperformance of the experiment in order to reduce the amount of multiple scattering andparticle interactions with material. An accurate description of the material is also requiredfor the simulation of the experiment and for estimating the amount of multiple scatteringthat particles undergo when performing track reconstruction. The material description isimplemented using the basic volumes in Geant4 [18] and there are limits on the accuracythat is achievable, both from the range of basic volumes (and combinations thereof) thatare available and from the CPU time taken to process very complex composite volumes.The appropriate balance must therefore be obtained between accuracy and simplicity foreach element of the description. This is particularly an issue for the description of thecomplex RF foil shape (see Fig. 6 below).
A comparison has been made between the measured masses of the various elementsof the detector, as determined at production, and their simulated counterparts, which in
10
general shows very good agreement, e.g. the average mass of the sensors is (2.17± 0.03) g,compared with the simulated mass of 2.14 g. Similarly the mass of the dominant componentof the material, the RF foil, is reproduced to within 2%. The RF foil was manufacturedas a single sheet of aluminium magnesium alloy and then pressed into its rather complexshape. So while the agreement of the total material may be good, the distribution of thatmaterial might not be so well described. Tests were performed by measuring the thicknessof an RF foil and it was found that the amount of material was too low in regions aroundthe beam in the description used up till 2012 and this has now been improved. While thishas only a small effect on the total material it does contribute significantly to the materialbefore the first measured point, which is important for the impact parameter performancewith lower momentum tracks (see Sect. 5.4).
The simulated detector description can be used to estimate the material traversed by aparticle as it moves through the VELO. Figure 5 (left) shows the amount of material alongthe trajectory of a particle, which originates from the interaction point, before it reachesthe radius of the first active strip on the detector (8.2 mm), as a function of pseudorapidityand azimuthal angle. The amount of material is expressed in terms of the fraction of aradiation length (X0) and the average material in the VELO acceptance passed throughbefore reaching this radius is 0.042X0. The average amount of material traversed before aparticle leaves the VELO at z = 835 mm is 0.227X0. The breakdown of this total materialbudget into the components of the VELO system inside the acceptance is shown in Fig. 5(right). The VELO design means that much of the electronics and services are outsidethe detector acceptance. The RF foil dominates with ∼ 43% of the material, and thenext largest component is the active silicon sensors. Considering the uncertainty on allcomponents of the VELO, the material traversed by a typical particle is estimated asbeing known to an accuracy of ±6%.
A particle traversing the VELO can interact with the material, potentially producing anumber of other particles in the interaction. If some or all of these particles are charged andtheir trajectories are within the VELO acceptance they can be tracked and the vertex ofthe interaction reconstructed. By examining the density distribution of these reconstructedvertices the material distribution can be studied. The procedure is applied to a sample ofdata arising from interactions of the beam with gas molecules in the beam pipe as thisprovides a more uniformly distributed flux of interactions than in collision events.
Figure 6 shows the distribution of vertices in a cross-section of the VELO betweeny = −5 mm and y = +5 mm. The vertices are plotted in the r′z plane, where r′ is theradial distance from the beam axis of the vertex multiplied by either 1 or −1 depending onthe sign of the x coordinate to separate the left and right halves of the detector. Verticeswith a radius less than 5 mm are not plotted. The top right and bottom left componentsof the figure focus on a more restricted region in z, just downstream of the interactionpoint in data and simulation. The main features of the R and Φ sensors in each moduleand the undulating form of the RF foil can clearly be seen. The most striking differencebetween data and simulation is the more angular form of the simulated RF foil.
Another useful feature of these vertex distributions is that they allow checks to beperformed of the relative positions of the various VELO components. The proximity of
11
]° [φ-100 0 100
η
1.5
2
2.5
3
3.5
4
4.5
5 0X
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.40XLHCb VELO average = 0.042
42.6%R.F. foil
Sensors 20.3%
14.1%Hybrids
6.6%Detector Supports
R.F. Box 5.8%
Kapton Cables 2.8%
Constraint System 2.5%
Connectors 2.3%
1.8%Paddle+Base
0.9%W.F. Suppressor
0.4%Cooling Block
LHCb VELO
0Total material = 0.227 X
Figure 5: (left) Material passed through on trajectories originating from the interaction pointand terminating at the radius of the first active strip in the detector expressed in terms of thefraction of a radiation length (X0) and as a function of pseudorapidity (η) and azimuthal angle(φ). The material is higher in the regions around ±90◦ where the two halves of the detectoroverlap. (right) Breakdown of the total material budget by component of the VELO. Thenumber given for each component is the percentage of the total average VELO material budget(0.227X0).
the sensors to the RF foil is of particular interest since if these were to touch it couldlead to an electrical short or damage to the RF foil. Figure 6 (bottom right) shows thematerial interactions in the region around one of the pile-up veto stations which have asingle sensor. It is clear that in this pile-up station the sensor is quite close to, though notactually touching, the foil. All of the VELO stations have a greater clearance from the foil.
These data can also be used to make a determination of the aperture available to thebeam due to the mechanical tolerances of the RF foil construction and positioning. Thisinformation is of particular interest for the upgrade of the VELO in which the radiusof the detector, and hence of the RF foil, is to be reduced [19]. The vertices attributedto interactions in the RF foil in the z regions around the sensors where the foil is at itssmallest radius are selected. Information can be obtained on the available beam apertureby fitting the data with a circle and extracting the variations [20]. Conservatively assuminga foil thickness of 300µm, an aperture of 4.9 mm is obtained, to be compared with thenominal value of 5.5 mm. This is further reduced to 4.5 mm by the weld of the foil to theRF box but is still well within the tolerance of 2.4 mm that was reserved for mechanicalimperfections of the foil with regard to the beam aperture.
12
Figure 6: Vertices of hadronic interactions in the LHCb VELO material. (top left) Full VELOsystem with entrance and exit regions visible in data. (top right) Zoom in to a group of sensorsdownstream from the interaction point in data. (bottom left) the same region reconstructedusing simulation events. (bottom right) Zoom onto an pile-up module consisting of a singleR-sensor to check the distance between the sensor and foil.
13
3 Calibration, monitoring and simulation
The performance of the VELO depends critically on its calibration. This section describesthe calibration of the timing and gain of the detector, and the data processing algorithmsthat are executed in FPGAs. The determination and verification of the parameters ofthese algorithms are discussed. The section starts with a brief description of the dataacquisition system (DAQ). The simulation of the detector is also described at the end ofthe section.
3.1 Data acquisition system
The silicon sensor’s strips are connected to the Beetle front-end ASICs [9] on the hybridcircuit boards of the VELO modules. The ASIC samples data with the LHC bunch crossingfrequency and stores them in an analogue pipeline, with a length of 160 events, awaitingthe decision of the first level trigger. Both the shape of the Beetle front-end amplifierpulse response and the sampling time can be tuned in order to get the best performance;the tuning procedure for the sampling time is described in Sect. 3.2.3.
After a positive trigger decision is obtained, the data from the Beetle ASIC are readout on analogue links [21]. Each 128 channel Beetle ASIC provides four-output ports.The signals of 32 channels of the sensor and four channels of header information are sentserially at 40 MHz rate. The system has a maximum readout rate of 1.1 MHz. The dataare sent out from the hybrid circuit board of the module on kapton cables, via vacuumfeedthroughs to the repeater boards that are located outside the VELO tank. The analoguedata are then transmitted over a 60 m long differential link (twisted pair) to the TELL1DAQ boards. The TELL1 boards are located in the counting rooms outside the radiationzone.
The TELL1 DAQ boards digitise the analogue signals and process the data in FPGAs.The adjustment of the sampling phase is described in Sect. 3.2.1. The processing algorithmsand the tuning of their parameters are described in Sect. 3.3. The TELL1 boards outputthe processed data to the high level trigger system [5]. The primary output from theTELL1 boards is zero-suppressed (ZS) data. For monitoring and tuning purposes, a numberof special output data types can also be transmitted. The raw ADC values (non-zerosuppressed, NZS) are sent out at a low rate of approximately 1 Hz. The data formats alsocontain the header data from the Beetle ASICs. The headers are used for the gain anderror bank studies (described in Sects. 3.2.2 and 3.4 respectively).
3.2 Timing and gain
The ADC sampling time for the digitisation of the analogue data needs to be determined,as does the input gain of the digitisation boards. The ADC sampling time is set for eachlink to account, for example, for the slight differences in cable lengths. The gain in theADCs is particularly important for the uniformity of the noise and signal levels measuredsubsequently. Then, the timing of the pulse sampling in the Beetle ASIC with respect
14
to the LHC beam collisions needs to be setup. These three calibration procedures aredescribed below.
3.2.1 ADC sampling time
As described in Sect. 3.1, one triggered event of a VELO sensor is fully read out through64 analogue links, that each carry 36 analogue voltage levels spaced by 25 ns each. Thefirst four levels in this readout block are encoded header bits, which have their heightscalibrated to be slightly (∼ 30%) larger than one MIP. On the receiving end, the TELL1boards are equipped with cards that digitise the 36 consecutive levels, sampling every 25 ns.The sampling can be delayed by several clock cycles (25 ns) and fine time adjustmentsmade with steps 1/16th of a clock cycle. The optimisation procedure first consists ofroughly aligning the readout of the front-end signals to the digitising window. This isperformed by sending internally generated signals, test-pulses, on known channels in theBeetle ASICs and adjusting the sampling so that the header bits are obtained in thecorrect positions. A scan is then performed over the fine time steps in order to get the finalsetting for each analogue link. The analysis of the scan data finds the optimal samplingpoint based on two conditions: the best signal to noise ratio and the minimal inter-symbolcross-talk. The inter-symbol cross-talk is defined as the fraction of the signal spilling overto the channel transmitted in the previous or next clock cycle on the same analogue link.A plot of the delay scan analysis is shown in Fig. 7, where the measured signal at eachtiming step is shown. The selected sampling point of the channels is indicated by the solidvertical line, while the small inter-symbol cross-talk at the sampling point of the previousand next channels is indicated with the dashed lines.
3.2.2 Gain
Each strip has a different capacitance, depending on its size and state of depletion, andthe length of its routing line. This, in addition to the range of variances in the Beetlepipeline, cause links to have a range of signal sizes for a given deposited charge. The gainis normalised with a method that is independent of these variations and is stable as theVELO undergoes radiation damage. Here, gain refers to the conversion factor that relatesa certain Beetle output voltage to an ADC value. The gain factor can be varied using ahardware setting on each digitisation card. This setting controls the upper limit of thevoltage range digitised by the ADC and hence the proportionality of an input voltage toADC counts.
The four bits of header information, which precede the 32 strip signals output on eachlink, encode the pipeline column number of the event. The headers are added to the linkoutput in the Beetle, so they are not affected by strip capacitance or by non-uniformitiesin the Beetle pipeline. They are digital bits, which are subsequently sent over an analoguelink, so can act as standard candles during the gain normalisation.
Depending on whether these digital bits corresponds to a ‘0’ or ‘1’, a header is classifiedas ‘Header High’ or ‘Header Low’, which when uncalibrated, typically have output valuesof around 560 and 460 ADC counts respectively. The distributions of high and low headers
15
Figure 7: An example of the scan over the digitisation phases (see text). Data for 16 analoguelinks are shown. The horizontal axis represents the sampling time, where the integer numbersindicate the location of the serially transmitted Beetle channels, and between two channels thereare 16 bins. The solid red vertical line indicates the chosen sampling point and the dashed redlines the position of the previous and next samples.
for each link are Gaussian in shape, and we define the full header swing (FHS) as thedifference between the two means. To calibrate the VELO, the gain of each link is set to avalue where FHS is equal to 100 ADC counts. The effect of a gain calibration on the FHSis shown in Fig. 8.
With the gain normalised, the most probable value of the Landau distributions andaverage noise values of each link are more uniform than for the uncalibrated VELO. Thegain has been found to be fairly stable and the calibration procedure is repeated everysix months or after the replacement of TELL1 digitisation cards or changes to the FPGAfirmware.
3.2.3 Timing to the beam
After performing the digitisation and gain determination, a tuning is required to synchronisethe front-end amplifier sampling to the signals generated in the silicon by particles frombeam collisions. The signal left by a particle passing through a strip is pre-amplified andshaped in the Beetle ASIC. The level of this pulse is sampled every 25 ns and stored in apipeline position. When a trigger command is received the Beetle outputs the data fromthe amplified signals that were sampled 4µs earlier; this trigger latency corresponds to160 clock cycles. The time alignment adjusts the Beetle sampling time to the time thatthe particles pass through the detector, which is synchronous to the LHC bunch crossingtime plus the time of flight of the particle to that sensor.
The time alignment is performed by scanning the clocks of the front-end readout insteps of 1 ns. Data from a few thousand collision events are stored for each step and the
16
80 90 100 1100
200
400
600
800
1000
1200
1400Calibrated
Uncalibrated
FHS (ADC counts)
LHCbVELO
Figure 8: The FHS distribution before and after the first gain calibration. The data were takentwo months apart in late 2009. One entry is made in the histogram for each link.
pulse-shape is reconstructed offline. The pulse-shape is fitted with a bifurcated Gaussianfunction, where the two halves have different widths but are constrained to have the sameamplitude at the peak; this accounts for the difference in the rise and fall time of thepulses. This simple function does not fully describe the peak or the tails of the distribution,most notably the undershoot after the pulse, but gives good agreement in the rising andfalling edges, and hence allows a quick and reliable way to optimise the sampling time.Conventionally the sampling time would be set to sample on the peak of the distribution,which would maximise the signal to noise distribution. However, we choose instead tominimise the contamination into the previous (pre-spill) or next (spillover) bunch crossing.This is obtained by setting a sampling time that equalises the spillover and pre-spill. Asthe rise time of the pulse is faster than the fall time this corresponds to sampling a fewnanoseconds after the peak, and corresponds to a loss of about 4% in the optimal signalsize. Figure 9 shows the result of a scan, with the simple function that is fitted shown andthe optimal sampling time marked.
3.3 FPGA data processing algorithms
After digitisation the data are processed in a series of FPGA algorithms performed onthe TELL1 board. Each TELL1 processes the output from one sensor, and contains fourFPGAs each of which independently processes 512 channels of data.
A full bit-perfect emulation of the FPGA processing algorithms was implemented inC-code at the same time as the VHDL code for the FPGAs was developed. Extensivechecks of the bit-perfectness are performed before any firmware changes, running boththe TELL1 and C-code software on the same data samples and ensuring that bit-perfectagreement is obtained. The C-code emulation is then encapsulated, in a software projectknown as Vetra [22] used by both the VELO and LHCb Silicon Tracker, and run inside the
17
Shifted sample time (ns)
-40 -20 0 20 40 60 80 100
Pu
lse
hei
gh
t (A
DC
co
un
ts)
-50
0
50
100
150
410
510
610
Shifted sample time (ns)
-40 -20 0 20 40 60 80F
itte
d p
uls
e h
eig
ht
(AD
C c
ou
nts
)
0
10
20
30
40
LHCbVELO
LHCbVELO
Figure 9: Pulse-shape obtained by combining data from all sensors, with a 1 ns time step andafter applying for each sensor the time shift determined by the procedure. The plot on the leftshows the ADC spectra versus time. The plot on the right shows the most probable value foreach time bin obtained from a fit of a Landau convolved with a Gaussian function. A bifurcatedGaussian function is fitted to this distribution from which the rise and the fall widths are obtained.The fitted function also predicts the amount of pre-spill (25 ns before the chosen point) and thespillover (25 ns after), which are marked with red vertical dotted lines. The chosen samplingtime is also marked with a red vertical dotted line.
main software framework of the experiment. This allows the algorithms to be fully testedwith data before being deployed in operation for collision data. This also allowed thealgorithms to be fully tested and debugged during the commissioning period. In operation,the ability to fully emulate the processing steps allows the output of any processing stageto be monitored, which is used for assessing the data quality and fully understanding thesystem performance.
The algorithms described below require a set of tuneable parameters to be determinedand uploaded to the FPGAs. The parameter values used are stored in XML files andrecorded in a database. The parameters are primarily based on the analysis of NZSdata using the Vetra package. For example, the pedestal values are determined and thecluster thresholds are optimised. In total 550k parameters are stored in the XML files foroperation of the system.
18
0 50 100 150 200 250
Raw
AD
C
420
440
460
480
500
520
540
560
580
600
620
0
50
100
150
200
250
300
350
Channel Number
LHCbVELO
Figure 10: Typical raw (Non-Zero Suppressed) data ADC values. Data from two ASICs, eachwith four analogue links, is shown. Variations in the pedestal values at the ASIC, analogue link,and channel level are seen.
3.3.1 Pedestal subtraction
Significant variation in the average raw ADC values (pedestals) are observed for each BeetleASIC. Variations in pedestal from link-to-link and channel-to-channel for the raw NZSdata are also seen, as shown in Fig. 10. Hence, the pedestal values must be determined foreach channel of the system individually, and thus 180k 10-bit pedestal values are required.
The pedestal values are determined offline using NZS data collected when there are nocollisions. The values are then uploaded to the FPGAs, and are subtracted from the rawADC when processing each event. Operational experience has shown that the pedestalvalues remain relatively stable and a pedestal retuning is made once the pedestal subtractedADC values of more than a few percent of channels are more than 2 ADC counts awayfrom 0. A retuning typically happens every two months, and is usually performed duringtechnical stops of the LHC. A retuning is also required after the replacement of a TELL1digitisation card or a change to the firmware. In addition to the pedestal subtraction,channels in the system that are known to be faulty are masked at this stage of the dataprocessing.
3.3.2 Mean common mode suppression
The 128 channels of the sensor that share the same Beetle ASIC, and the 32 channelsof one analogue link may be subject to the same sources of signal fluctuation, known
19
-120 -100 -80 -60 -40 -20 0 20 40 60
Lin
ks x
Eve
nts
1
10
210
310
410
510
610
710
Common Mode Noise (ADC counts)
LHCbVELO
Figure 11: Distribution of common mode noise per link. The blue histogram is for all links. Thegreen histogram is for links in a ASIC that has large signals (see text).
as common mode (CM) noise. The CM noise is suppressed in the FPGAs with a MeanCommon Mode Suppression (MCMS) algorithm which operates on each analogue link,correcting any baseline offset in each event.
In the algorithm, first the average pedestal-subtracted ADC value of the channels ina link is calculated for that event. Using this mean value a search for particle hits isperformed for each channel. All channels with hits are masked, and a new mean value iscalculated for each link. This value is then used to correct the ADC values in all channelsof the link.
This algorithm would not fully correct common mode sources from the silicon sensor,for example from noise pickup from the proton beams. This is because consecutive channelson the sensor are not read out on consecutive inputs of the front-end ASIC (see Sect. 3.3.3).A second common mode algorithm was originally foreseen for use after channel reordering.However, studies performed by monitoring the noise level as a function of the VELOopening distance, showed that this source of noise pickup was small, and this secondcommon mode algorithm has not been used in operation.
Figure 11 shows the distribution of the CM noise measured in the system. It peaks at0 and has a sigma of 1.74 ADC counts. Furthermore, out of the 1.74 ADC counts linkCM noise, there is a common fluctuation among all 5376 links, introduced by the commonpower supply and environment. This global common mode noise is 1.47 ADC counts,meaning the intrinsic link common mode noise is 0.93 ADC counts. For comparison theaverage incoherent (CM suppressed) noise is 1.91 ADC counts.
The common mode distribution has a long tail on the negative side, where all four
20
1000 1050 1100 1150
500
550
600
650
700
750
Raw ADC
Pedestal
AD
C V
alue
Sensor 7, Run 93410, Event 2035787399
(a)
Channel ID
LHCbVELO
Charge Injected In One Chip [MIPs]0 50 100 150 200 250
Chi
p B
asel
ine
[AD
C C
ount
s]
-30
-25
-20
-15
-10
-5
0 Measurement
Fit, slope=-0.118(b) LHCb
VELO
Figure 12: (left) Event snapshot for an ASIC that has large signals (channels 1024-1152). Theblue dots are the raw ADC values. The red lines are the pedestals. (right) ASIC baseline shift inresponse to injected charge in units of the most probable charge deposited by a MIP in 300µmthick silicon.
links from the same ASIC have a coherent baseline shift. The effect is illustrated in anevent snapshot in Fig. 12 (left). Accompanying the baseline shift are a few channels withvery large signals that reach the saturation point of the electronics, most likely arisingfrom soft tracks at large angles. In Fig. 11 the green distribution corresponds to links thathave at least two channels with signal more than 180 ADC counts above the baseline. Thelarge energy deposition hypothesis was confirmed in a test-pulse calibration procedure. Acontrollable amount of charge is injected to a few selected Beetle ASIC channels usingtest-pulse calibration capacitors. The baseline shift is proportional to the injected chargeas shown in Fig. 12 (right). These baseline shifts are corrected by the CM correctionalgorithm.
3.3.3 Topological strip channel reordering
As a consequence of the layout of the routing of the channels from the VELO R and Φsensors on the second metal layer, neighbouring channels on the sensor are not necessarilyread out consecutively. In the Φ sensor strips from the inner region of the sensor are routedout over strips in the outer region, and the readout is thus intermingled between inner andouter region strips, as shown in Fig. 2. Therefore the electronic readout ASIC channelnumbering does not correspond to the physical (strips) channels on the sensors. Thezero-suppression and clusterisation must be performed in terms of physically neighbouringchannels, and thus a translation algorithm is necessary. The reordering procedure is
21
performed separately for R and Φ type sensors. The implementation of this algorithmconsumes a significant amount of the FPGA resources.
3.3.4 Zero suppression and clusterisation
The final data processing algorithm on the TELL1 produces the raw cluster bank which issent out to the high level trigger. This processing reduces the data size by only transmittingchannels that have significant signals present. It also groups together neighbouring hitsthat may be due to single particles and produces a first estimate of the centre of thiscluster.
In the first step, hit detection is performed by finding signals above seeding thresholds.The current seeding threshold in each channel is set to be six times higher than themeasured noise in that channel, rounded to the nearest ADC count. This limits fake noisehits while not significantly reducing the selection efficiency of true signal clusters in fullydepleted sensors (see Sect. 4.1).
In the second step, the algorithm attempts to include adjacent strips in the cluster.Strips are added to the cluster if they pass an inclusion threshold, which is currently setat 40% of the seeding threshold. By setting a lower inclusion threshold than the seedingthreshold additional charge is added to the cluster, which helps improve the resolution onthe particle position. The maximum number of strips per cluster is four.
The centre of the cluster is calculated from a pulse-height weighted average of thestrips contributing to the cluster. This is calculated with a precision of an eighth of astrip (3-bits). This calculation is used to save time in the trigger; it is recalculated withfloating point precision for use in the offline tracking. The resolution and cluster detectionefficiency are discussed in Sect. 4.
Finally, each cluster is encoded into a bit structure that contains information on thecluster position, cluster size, and the charge measured on each strip that contributes tothis cluster. The size of the raw cluster bank varies depending on the event’s occupancy,and the typical size of this bank for the full VELO is about 20 kB.
3.4 Error identification
Information is sent out from the Beetle in the header bits that allow a number of consistencychecks to be performed. Error banks are then produced by the TELL1s when the data ofa link is internally inconsistent, or there is inconsistency between data from various links.They aim to contain the required information to trace down the origin of the error and takeappropriate measures. The most commonly occurring errors relate to the verification ofthe pipeline column number. This number records in which column of the Beetle pipelinethe data are stored while awaiting a first level trigger acceptance signal. The pipelinecolumn number is reset periodically by a broadcast command and hence verifying that itis identical for all front-end ASICs ensures the synchronicity of the detector. Elementsof the binary value of the pipeline column number are transmitted as an analogue signalpreceding the data from each link of the ASICs. Gain variations and inter-symbol cross-
22
talk can cause the analogue values to end up in between the corresponding low and highbit thresholds, and consequently make the bit assignment ambiguous. Calibration andverification procedures that are executed regularly, have managed to control these effectsto a satisfactory level and errors now occur at the level of a few in every ten thousandevents.
3.4.1 Single event upsets
One error type of particular interest is a single event upset (SEU). Since the Beetle ASICis exposed to ionising radiation charge will be liberated in its electronic circuits. Mostparticles will however not generate enough charge to influence the operation of the devices.Occasionally a highly ionising particle can cause a change in the logic state of one of theflip-flops in the Beetle ASIC. These SEU can affect the performance on the ASIC if theyoccur for instance in one of the configuration registers. Therefore all critical flip-flops areimplemented in a triple redundant way, featuring majority voting and auto-correction. Tomonitor the rate of SEU, all bit-flips are counted in a dedicated counter. The two leastsignificant bits of this counter are sent with the data for every trigger as part of the headerbits. By analysing the transitions of these counter bits, an average of 2.9 SEU per pb−1
of delivered integrated luminosity for all 1344 Beetle ASICs combined is observed.
3.5 Monitoring
The VELO monitoring infrastructure is used to ensure that any degradation in dataquality is observed and can be followed up quickly. It is complementary to the LHCbmonitoring which is split in two components: online and offline monitoring. The LHCbonline monitoring analyses a fraction of the ZS data sample that has been selected by thetrigger. The LHCb offline monitoring uses a number of lightweight algorithms to monitor asample of all data that are processed for physics analyses. The VELO monitoring adds twocomponents: an analysis of NZS data and a detailed analysis of the ZS data. Experienceshows that many data quality problems are spotted first in the detailed checks performedby the dedicated VELO monitoring team.
The NZS data are recorded in a special data stream which is stored separately from thephysics data at a rate of 1 Hz. Events directed to this stream contain NZS data from allVELO sensors. This allows the analysis of pedestals, common mode suppression, cross-talkeffects and noise to be made for each run. The NZS readout has also been used to studycorrelations between different sensors in the same event (see Sect. 3.3.2).
The VELO monitoring also analyses ZS data which have not been biased by any triggerselection. This sample of events is thus representative of the data being recorded by thedetector, and allows the detector performance to be monitored independent of changesto the trigger. These data are recorded at a rate of 10 Hz and stored in a separate datastream to increase the efficiency of the data analysis.
All data in the ZS and NZS streams used by the VELO monitoring are processedautomatically by monitoring algorithms shortly after they have been recorded. The output
23
Figure 13: Example of the graphical user interface used in the VELO monitoring. Shown aredistributions of the number of hits per VELO track segment (top left), the track angle withrespect to the beam axis (top right), and the ADC counts of R and Φ sensor clusters associatedto a track (bottom left and right, respectively). The results of the current run are shown in redfilled symbols and the reference data for comparison are shown as black open symbols.
of the monitoring algorithms are analysed daily by VELO data quality shifters. These arepeople on call who analyse the incoming data and alert experts of problems if necessary.These analyses include information on pedestals, noise, common mode, cluster signaldistributions, occupancy, tracking and alignment. Separate monitoring is implemented forall slow control data such as voltages, currents, temperatures and pressures.
A graphical user interface has been developed to facilitate the analysis of the output ofthe regular monitoring algorithms. An example of its usage is shown in Fig. 13. The sametool also produces trend and correlation plots of the relevant quantities, which are used totrack the detector stability over time. Summary tables of representative values to assessdata quality, for example numbers of high occupancy channels, for each run are also made.The summary values and comments from the analyst of the data are stored in a dedicatedelectronic logbook.
Data are also taken under special conditions to cover aspects that cannot be easilyaccessed with physics data. Dedicated monitoring is in place for these periodic performancechecks. These include sensor current versus voltage, current versus temperature, noise
24
versus voltage, and charge collection efficiency versus voltage data which are performed tomonitor radiation damage (see Sect. 4.6).
3.6 Simulation
The LHCb simulation is based on Monte-Carlo event generators and the use of the Geant4toolkit [18]. A detailed model of the detector material in the VELO has been produced,through which the simulated particles are propagated (see Sect. 2.6). The entry and exitpoints of the simulated particles in the silicon sensors are obtained from Geant4. A thinsilicon sensor is expected to have an energy deposition distribution that can be describedby a Landau function convolved with a Gaussian distribution. The charge deposited bya particle in the silicon sensor is calculated using the form from theory and previousexperiment [23].
The total deposited charge is distributed at uniformly spaced nodes along the particlepath in the silicon, with the energy fluctuations along the path simulated. The diffusionof the charge as it drifts in the electric field of the silicon is simulated through a Gaussiansmearing whose width is dependent on a bias voltage parameter. Cross-talk between thecharge attributed to the strips is then added to model capacitive coupling. Noise is addedto the strips following the measured values in the data. The measured pulse-shape of thefront-end electronics is also simulated, and corrections are applied for the time of flightof the particles. The electronics response from the previous and the next events are alsosimulated by applying the pulse-shape to simulated events. Pedestal offsets and commonmode noise are not normally simulated as they are removed effectively in the real data bythe FPGA processing algorithms. This simulation stage results in a model of the amountof charge collected on a sensor’s strips. The model has the bias voltage and capacitivecoupling as free parameters which are tuned to obtain agreement with the measured dataresolution as a function of pitch and track angle and typically agree to within 5%. Thissimple algorithm is fast and sufficiently accurate for the simulation required for physicsanalyses.
The next stage of the simulation models the processing in the TELL1 board. First,the signal is digitised. At this point the DAQ emulation produces an output format thatcorresponds to the raw non-zero suppressed data. The simulated data are then passedthrough the bit-perfect emulation of the TELL1 clusterisation algorithm (see Sect. 3.3).Identical clusterisation thresholds are used to those in the data taking. The output databank is then produced in the same format as that produced by the TELL1 boards. Aglobal scaling, the gain (electron to ADC conversion), is applied to normalise the observedsignal in simulation, with a correction for the variation in signal size with radius (seeSect. 4.1). The trigger, pattern recognition and track reconstruction algorithms are appliedidentically for real and simulated data.
Sensor 40 (R) : Red pointsSensor 104 (Phi) : Blue circles
Figure 14: (left) Fit of the signal distribution for Sensor 104 (a Φ sensor) for clusters on longtracks with a track intercept point on the sensor 10 < r( mm) < 15. (right) The most probablevalue of the fits to sensors 40 and 104, the R and Φ sensor of the same module respectively, as afunction of the track intercept point radius. The cluster ADC counts were normalised to a trackcrossing 300µm of silicon.
4 Overall system performance
This section characterises the VELO system performance, including the signal to noise ratio,hit resolution, occupancy, and efficiency studies. Beam backgrounds are also discussed.The results presented use data before appreciable effects of radiation damage were observed,results on the observed radiation damage are reported at the end of this section.
4.1 Signal size and noise rate
The observed total signal from the clusters is determined by fitting a Landau convolvedwith a Gaussian function around the peak region. The deposited charge for each trackis corrected for the track’s angle to give a result corresponding to a path length in thesilicon of the nominal 300µm thickness of the sensor. An example fit is shown in Fig. 14(left). The function provides a good description of the rising edge and peak region, butundershoots the data in the high energy deposition region. This is in part due to thepresence of photon conversions to electron positron pairs which, in the absence of anappreciable magnetic field in the VELO, can remain merged into a single cluster. Thetuned simulation is in good agreement with the data distribution, including in the highenergy deposition region, and the FWHM does not require any additional smearing of thedetector response. The most probable value (MPV) of the Landau distribution varies as afunction of track radius, as shown in Fig. 14 (right).
The noise of the front-end ASIC depends on the strip capacitance. On the R sensorsthe inner radius strips have the lowest capacitance due to their shorter length though thisis partially compensated due to the inner strips having longer routing lines. Still, the
26
R strip number0 500 1000 1500 2000
Aver
age
nois
e in
AD
C c
ount
s
0
0.5
1
1.5
2
2.5
3
LHCb VELO
Phi strip number0 500 1000 1500 2000
Ave
rage
noi
se in
AD
C c
ount
s
0
0.5
1
1.5
2
2.5
3
LHCb VELO
Inner Phi strips Outer Phi strips, with routing line
Outer Phi strips, no routing line
Figure 15: Noise in ADC counts averaged across the 42 installed R (left) and Φ (right) sensors,with the error bars indicating the RMS of the distribution.
dependence of the noise on the strip length is visible when comparing Fig. 15 (left) withthe sensor layout shown in Fig 2. The R sensor is divided into four approximately 45◦
segments, and the strip length increases with increasing strip number in each segment.The Φ sensor has two zones with inner and outer strips. The inner strips are shorter buthave additional routing line contributions to their capacitance. In the outer zone everyalternate strip is under the routing line for an inner strip so the capacitance for thesestrips is larger. The noise in these three types of Φ sensor strips is shown in Fig. 15 (right).Larger noise is also clearly visible in both the R and Φ sensors every 32 channels, this isdue to inter-symbol cross-talk from the digital header information into the first channel ineach analogue readout link. A suppression algorithm for this inter-symbol cross-talk hasbeen implemented in the FPGAs, but is not currently used due to the small size of thiscross-talk and the large signal to noise ratio.
The average signal to noise ratio, computed as the average MPV of single strip clustersdivided by their strip noise, for the VELO is around 20:1. It is higher for the Φ sensorsthan the R sensors and shows a variation on the sensor radius as shown in Fig. 16.
4.2 Resolution
The hit resolution in silicon devices depends on the inter-strip readout pitch and thecharge sharing between strips. The charge sharing varies with operational bias voltageand the projected angle of the track. The bias voltage was 150 V throughout the physicsdata taking in 2010–2012. The projected angle provides information on the number ofstrips that the particle crosses while it traverses the thickness of the silicon sensor. Itis defined as the angle between the track and the perpendicular to the sensor, in theplane perpendicular to the sensor and containing the perpendicular to the strip. Initiallythe resolution improves with increasing angle, due to the charge sharing between stripsallowing more accurate interpolation of the hit position. The optimal resolution is obtained
27
Radius [mm]10 15 20 25 30 35 40
S/N
for c
lust
ers
15
16
17
18
19
20
21
22
23
24
25
LHCb VELO
Sensor 40 (R) : Red pointsSensor 104 (Phi) : Blue circles
Figure 16: Signal to noise (S/N) ratio from the MPV of the signal for single strip clusters ontracks divided by the noise of that strip. Shown are the S/N values for sensor 40 (R) and sensor104, the Φ sensor of the same module, as a function of impact point radius.
when the tracks cross the width of one strip when traversing the 300µm thickness of thesensor. For the VELO the optimal projected angle varies between about 7◦ at the lowestinter-strip pitch of 40µm, to about 18◦ for the largest 100µm pitch strips. Above theoptimal angle the resolution begins to deteriorate due to the fluctuations in the charge onthe strips and because the signal to noise ratio on individual strips may drop below theclustering threshold.
The clustering algorithm and charge interpolation method is described in Sect. 3.3. TheVELO reads out analogue pulse-height information from the strips, and this information isused offline to calculate the cluster position using the weighted average of the strip ADCvalues. Including the track angle dependence in the clustering algorithm is found to give asmall improvement in precision. The results presented here rely on the offline recalculationof the position, while the trigger relies on the lower resolution (3-bits) calculation (seeSect. 3.3.4). The estimated resolution in the simulation is parameterised and fitted as afunction of both track angle and strip pitch. This resolution estimate for each hit is thenused in the Kalman fit tracking algorithm.
The hit resolution is determined from the hit residuals which are evaluated using theLHCb Kalman filter track fit [24] and include a correction for multiple scattering andenergy loss dependent on the track momentum. The residual is defined by the distancebetween the hit measurement and the extrapolated point of the fitted track to that sensor.As the hit for which the residual is being determined is included in the track fit this givesrise to a bias in the residual which must be corrected for. The bias correction used todetermine the residual is
√VM/VR [25] where VM is the variance of the measurement and
VR is the variance of the residual. The evaluation of this correction is implemented in theKalman fit.
0.45 mµpitch bin (40-45) mµpitch bin (50-55) mµpitch bin (75-80) mµpitch bin (95-100)
LHCbVELO
Figure 17: (left) The VELO resolution for two projected angle bins for the R sensors as afunction of the readout pitch compared with binary resolution. (right) Resolution dividedby pitch as function of the track projected angle for four different strip pitches.
The resolution has been determined as a function of the strip pitch and of the projectedangle. For each bin, the resolution has been determined from the sigma of the fit ofa Gaussian function to the distribution of the corrected residuals. The resolution isevaluated using tracks that have hits in the tracking stations behind the magnet andhence for which the momentum measurement is available. The tracks are required to havea momentum greater than 10 GeV/c to reduce the dependence on the estimation of themultiple scattering effect, and a number of other track quality criteria are applied to rejectfake tracks. The results are presented here for the R sensor. The Φ sensor results arecompatible but the almost radial geometry of the strips means that tracks primarily havesmall projected angles.
The measured hit resolution has a linear dependence on the strip pitch in projectedangle bins, as shown in Fig. 17 (left). The hit resolution at small projected angles, almostperpendicular to the sensor, has a resolution which is close to that which would be obtainedfrom a binary system. This is to be expected as the charge sharing between strips at thisangle is minimal. A significantly better resolution is obtained for larger projected angles,where the fraction of two strip clusters increases and the analogue readout of the pulseheight in each strip is of benefit. The hit resolution as function of the projected angle isshown in Fig. 17 (right) and the fraction of one and two strip clusters as a function of theprojected angle and strip pitch are shown in Fig. 18. The best hit precision measured isaround 4µm for an optimal projected angle of 8◦ and the minimum pitch of 40µm.
29
Projected Angle (degree)0 5 10 15 20
Perc
enta
ge 1
str
ip c
lust
er [%
]
0
20
40
60
80
100
120 mµpitch bin (40-45) mµpitch bin (50-55) mµpitch bin (75-80) mµpitch bin (95-100)
LHCbVELO
Projected Angle (degree)0 5 10 15 20
Perc
enta
ge 2
str
ip c
lust
er [%
]
0
20
40
60
80
100
120 mµpitch bin (40-45) mµpitch bin (50-55) mµpitch bin (75-80) mµpitch bin (95-100)
LHCbVELO
Figure 18: The percentage of one (left) and two (right) strip clusters as a function of the trackprojected angle for four different strip pitches.
4.3 Occupancy
The detector occupancy is a key parameter in the performance of the pattern recognition andtracking algorithms of the experiment. High occupancy can lead to the mis-identificationof hits on tracks and increase the number of hit combinations decreasing the speed ofthe algorithms. The occupancy shown here is for clusters. The cluster seeding thresholdsand masking of noisy strips in the FPGA data processing algorithms ensure that thecontribution to the occupancy from noise is negligible compared with that from particles;in the absence of circulating beams the observed occupancy is below 0.01%. The typicalcluster occupancy during 2011 operations is shown in Fig. 19 (top plots). Only eventsfrom particle beam crossings are utilised in the computation, and this data sample hasan average number of visible interactions per beam crossing, µ, of 1.7. The occupancy isshown computed with data collected using two different triggers. Data from a randomtrigger on beam crossings are used as this represents the average occupancy in the eventsobserved by the detector. The occupancy for events passing the high level trigger isalso given, this is higher as events with heavy flavour production are typically of highermultiplicity than the average. The distribution for events passing the high level trigger isnot fully symmetric around the collision point due to the preference of selecting events inthe LHCb acceptance.
The cluster occupancy has a dependence on the position of the sensors along thebeam-line, as shown in Fig. 19 (left). The location of the interaction region is clearlyvisible from the dip in the occupancy distribution. The highest occupancy of 1.1% forevents passing the trigger is at around the end of the closely spaced region of sensors inthe VELO, where the occupancy is 44% larger than at its minimum.
The occupancy also varies across the sensor, increasing closer to the beam as seen
30
200 0 200 400 600Z [mm]
0.2
0.4
0.6
0.8
1.0
1.2R
Sens
or O
ccup
ancy
[%]
HLT random
40 20 0 20 40Rlocal [mm]
0.2
0.4
0.6
0.8
1.0
1.2
R Se
nsor
Occ
upan
cy [%
]
HLT random
LHCb VELO
200 0 200 400 600Z [mm]
0.1
0.2
0.3
0.4
0.5
0.6
0.7
R S
en
sor
Occ
up
an
cy [
%]
0 mm 1 mm 5 mm10 mm20 mm29 mm
40 20 0 20 40Rlocal [mm]
0.1
0.2
0.3
0.4
0.5
0.6
0.7
R S
en
sor
Occ
up
an
cy [
%]
0 mm 1 mm 5 mm10 mm20 mm29 mm
Figure 19: Cluster occupancy in the VELO silicon strip detectors. (left plots) Average occupancyin the sensor as a function of the position of the sensor along the beam-line. (right plots)Occupancy as a function of the local radius of strips on the R sensors, with a negative signapplied to sensors on one half. The upper plots show the occupancy when fully closed using 2011data with a µ of 1.7 and for events selected using a random trigger or with events passing thehigh level trigger. The lower plots show the occupancy as a function of closing distance withfully closed labelled as 0 mm, the points below then follow in order of the retraction distanceindicated in the key.
in Fig. 19 (top right). The maximum occupancy is 83% higher than the minimum asa function of R. This is a much weaker dependence than might naively be expectedsince the inner strips are five times closer to the beam than the outer strips. However,
31
the semi-circular sensor geometry means that the strip length naturally reduces withdecreasing radius on the R sensors, and the inter-strip pitch also reduces with decreasingradius on both the R and Φ sensors. The occupancy reduces in the smallest radius regionon the R sensors as these strips are shorter due to having the corners cut off.
Figure 19 (lower plots) also shows how the occupancy varies across the sensor at a rangeof closing distances. These data do not have a well defined µ as the luminosity levelling(see Ref. [26]) had not yet been performed. The minimum average sensor occupancy dropsfrom around 0.4% when closed to around 0.1% when the VELO sensors are fully retractedby 29 mm from the beam. Even in the fully retracted position the number of particlesthat can be reconstructed in the detector allows the performance of the system to bedetermined and the primary vertex to be fitted to allow the detector to be safely closed(see Sect. 2.5).
4.4 Beam backgrounds and high multiplicity events
Studies have been performed of beam-related backgrounds in the VELO, and a number ofsources of hits have been identified: beam-gas interactions; instrumental effects triggeredby the presence of beam; and beam interactions with collimators. The rate of these effectsis sufficiently low to not have a significant detrimental effect on physics or to contributesignificantly to radiation damage in the detector.
The interaction of beams with residual gas in the LHC vacuum pipe and VELO vacuumvessel provides a useful data sample for alignment and luminosity studies. Simulations ofbeam-gas interactions in the VELO vacuum vessel and the long straight section of theLHC have been performed [27], with the latter including particle fluxes from beam-gasinteractions in the long straight section and from proton interactions in collimators. Tracksarising from these beam-gas interactions provide a complementary sample to tracks frombeam-collisions due to their very forward angular distribution. Many of these tracks passthrough all VELO sensors and this provides useful additional constraints for alignment(see Sect. 5.2). The reconstruction of the beam-gas interaction vertices in the VELOregion is illustrated in Fig. 20 (left). The blue and red points are obtained from eventsin which only one beam passed through LHCb, and vertices with at least five tracks arereconstructed from interactions of the beam with residual gas. The horizontal crossingangle of the beams is apparent. The downstream beam (red) interactions are limited intheir extent in the negative direction along the beam-line by the predominantly forwardacceptance of LHCb. This reconstruction of beam-gas interactions allows a measurementof the transverse bunch profile along the beam trajectory to be obtained and allows thebeam angles, profiles and relative positions to be determined. The beam-overlap integralcan then be obtained from the extracted beam profiles and has been used to obtain ameasurement of the luminosity [28]. The precision of the VELO allows this luminositydetermination to be performed with a comparable precision to that obtained from thewell-known van der Meer scan method. The rate of beam-gas interactions can also beincreased in the VELO by the use of a gas injection system, which was commissioned in2011 [29].
32
LHCbVELO
X [
]
Z [ ]
Figure 20: (left) Imaging of the LHC beams through the reconstruction of the production vertexof tracks in 2010. The vertices from beam-gas interactions with the beam travelling in the +z(−z) direction in LHCb are shown in blue (red). Vertices arising from beam-beam interactionsare in green. The vertical axis represents the horizontal direction (x) and the horizontal axis thebeam direction (z). (right) An event display image showing clusters in R and Φ sensors with ahigh occupancy in a localised region assumed to be due to a beam background splash event.
Events containing a collimated spray of background hits are also observed, as shownin Fig. 20 (right). These events are characterised by high occupancy in local regions ofthe sensor, and they are correlated with activity in other tracking detectors. The rate ofevents is proportional to the beam intensity in the LHC, and typically constituted 0.03%of events passing the high level trigger in 2011. A possible origin for these events is beaminteractions with collimators.
4.5 Efficiency and faulty channel analysis
Studies have been performed of the charge collection efficiency and cluster finding efficiencyof the detector. The detector is initially operated at a bias voltage sufficient to obtainthe full charge from the sensors, and the charge collection efficiency and cluster findingefficiency as a function of radiation are reported in Sect. 4.6. In addition faulty channelshave been identified through the study of the cluster finding efficiency, occupancy andnoise spectrum.
The cluster finding efficiency is a useful measure of the performance of the detector.This is determined by excluding a sensor in the pattern recognition, and interpolatingtracks to this sensor. The tracks are required to have hits in both the R and Φ sensorsin the two modules before the test sensor and in the two modules after the test sensor.These requirements place some restrictions on the sensors that can be probed and theregions analysed within the studied sensors. Tracks that have intercept points inside theactive region of the sensor are then considered. Track quality selection cuts are applied toremove fake tracks and track isolation cuts to prevent a cluster from another track beingselected. The efficiency with which clusters are found near the track intercept points isthen determined. The efficiency for finding a cluster depends on the applied bias voltageand the thresholds applied in the cluster making algorithms. The standard operational
33
Module number5 10 15 20 25 30 35
Clu
ster
Fin
ding
Effi
cien
cy [
% ]
90
92
94
96
98
100
R-Sensor
Phi-Sensor
All Tracks
All strips-1 = 0 pbL LHCb VELO
Module number5 10 15 20 25 30 35
Clu
ster
Fin
ding
Effi
cien
cy [
% ]
99
99.2
99.4
99.6
99.8
100
R-Sensor
Phi-Sensor
Bad Strips Excluded
Bad strips excluded-1 = 0 pbL LHCb VELO
Figure 21: The cluster finding efficiency for each sensor with all strips (left) and when identifiedbad strips are excluded (right).
settings are used for the studies presented here.The region boundary between the inner and outer strips of the Φ sensor and the middle
region boundary of the R sensor both contain a 79µm wide region with the HV biasresistors. This region and the region of the guard ring around the sensor are excluded inthe efficiency calculation. Averaging across sensors, cluster finding efficiencies of 99.45%sensors were obtained at the start of operation as shown in Fig. 21. Bad strips are identifiedwith the cluster finding efficiency analysis described below. When these bad strips areneglected the cluster finding efficiency rises to 99.97%. The system contains only a singlefront-end ASIC that is not functioning, the failure of which occurred after production andbefore the start of physics operations, this is visible in the reduced efficiency of the Φsensor of module 21 in Fig. 21 (left).
During the production of the modules faulty channels were identified through theapplication of three techniques: high resolution visual microscope inspections, the responseof the modules to scans of a laser, and the analysis of noise data. A total of only 0.6% ofthe VELO sensor strips were found to be faulty, see Table 1. The analysis of noise datacan also be performed on the installed experiment. The measured noise on a functioningchannel will decrease once it is biased above its depletion voltage, since its capacitancewill decrease: unbonded strips or faulty ASIC channels will not display this dependence.However, it was found that the tuning of the selection was highly detector and conditionsensitive and hence this method of faulty channel identification is less accurate for repeatedsemi-automatic analysis than those discussed below.
The cluster finding efficiency analysis can also be used to determine the number of deadchannels in the detector. A track is extrapolated to the detector and clusters searchedfor in the channels around the intercept point. A channel is identified as dead if tracksextrapolated to this strip show four times more missed clusters than the mean of the 30
34
Table 1: Fraction of faulty strips classified as dead or noisy. The results are obtained fromthe methods used at production, occupancy spectrum studies, and a cluster finding efficiencyanalysis. Values are given at the time of production or start of operations in 2010, at the end of2011, and at the end of 2012.
Figure 22: The intercept position of an extrapolated track when a cluster is not found for twoexample R (left) and Φ (right) sensors. The bad channels are clearly visible.
nearest neighbours. Examples of the locations of strips in which clusters are not found areshown in Fig. 22, which shows the boundary region on the Φ sensor between the inner andouter strips and the location of bad strips. This method identifies 89% of the channelsfound in the dead channel list made at production. Only 0.8% of channels were identifiedas dead at the start of operations by the cluster finding analysis method. At the end of2011, after two years of operation and a delivered fluence of 1.2 fb−1, the fraction hadincreased to 1.1% of channels.
Two classes of faulty strips, dead and noisy, are also identified through the analysis ofthe channel occupancy spectra (see Sect 4.3). Faulty channels are usually masked in theFPGA data processing algorithms, and hence would not show up in the occupancy spectra.Therefore, the analysis is performed using NZS data and emulating the data processing,
35
but without the faulty channel mask applied. Channels with less than 10% of the averageoccupancy are identified as dead, and those with greater than three times the averageoccupancy as noisy. The results are compared with those found at production and withthe cluster finding efficiency method in Table 1, and seen to be in good agreement. Thelevel of dead and noisy strips is shown to have remained at a low level throughout theoperations.
4.6 Radiation damage studies
The proximity of the silicon sensors to the LHC beam results in a high particle fluence ofup to 5 × 1013 1 MeV neutron equivalents / cm2 ( neq) per fb−1 of delivered integratedluminosity for the most irradiated sensor regions, varying by approximately a factor of twoas a function of the position of the sensors along z. The VELO geometry, with the sensorsplaced perpendicular to the beam, gives rise to a highly non-uniform radiation dose acrossthe sensors with the fluence falling off with radius r as approximately r−1.9. The estimatedfluence is obtained from the LHCb simulation and using measured values of displacementdamage in silicon. Particle irradiation gives rise to both bulk and surface damage effects inthe silicon. The bulk radiation damage is primarily caused by the displacement of atoms inthe silicon sensors from their lattice sites, and induces changes in the leakage current andeffective doping concentration of the material. Consequently, radiation damage monitoringwas put in place from the start of operations. The leakage currents, noise and chargecollection efficiency (CCE) of the sensors are studied regularly, with dedicated data takingscan procedures having been developed. Detailed studies of radiation damage in the VELObased on 2010 and 2011 data taking have been reported in Ref. [30]. Further results, basedon the 2012 and 2013 data taking periods, are summarised here.
4.6.1 Current measurements
Sensor currents are studied as a function of both voltage and temperature. Current-voltagescans are taken with an automated procedure on a weekly basis. The shapes are analysedto look for signs of gradient changes that could indicate the onset of breakdown, andthe values at the operational voltage of 150 V are compared to the expected currents.Current-temperature scans are taken a few times a year by controlling the temperatureof the cooling system. The current-temperature scans allow the surface and bulk currentcomponents to be measured separately. More details of these studies are available inRefs. [31, 32].
The expected increase in the bulk current in each sensor due to radiation damage iscalculated using the predicted fluences and the measured temperature history of the sensorand the relation [33]
∆I = αφVSi,
where α is the annealing parameter in units of A/ cm, which depends on the temperaturehistory, φ is the fluence in particles per cm2 and VSi is the silicon volume in cm3. The
36
Figure 23: Currents measured for each sensor as a function of time (bottom). The integratedluminosity delivered to LHCb and the average sensor temperature is shown over the same timescale (middle and top). Increases in the delivered luminosity are matched by increases in thesensor currents. The evolution of the mean measured current agrees well with the predictionfrom simulation. The mean measured value excludes sensors that are surface-current dominated.
evolution of the observed currents in the sensors with delivered integrated luminosity arein good agreement with the expectation (see Fig. 23).
4.6.2 Effective doping concentration
The n-bulk sensors undergo space-charge sign inversion under irradiation, and hence theirdepletion voltage initially reduces with irradiation. This continues until type inversionoccurs, after which it increases with further irradiation. In order for the charge collectionefficiency of the sensors to remain reasonably high, the sensors must be close-to or fullydepleted during operation. As all of the VELO sensors are operated at a constant voltageover long periods, monitoring the sensor depletion voltages is a useful experimentaltechnique for ensuring that the CCE for a particular sensor does not decrease significantlydue to the sensor being under-depleted. In practice, this is achieved by monitoring theeffective depletion voltage (EDV), which is derived using the following method. Here wereport results for the n-type sensors, but note that one p-type module is also installed inthe VELO and is studied in Ref. [30].
37
Sensors are grouped into a 1–in–5 pattern, where four of the sensors are operatedat the nominal operation voltage (150 V throughout 2010–2013) whilst a single sensor,referred to as the test sensor, has a range of bias voltages applied to it. Using only thesensors at the nominal operation voltage, a track is then fitted and extrapolated to thetest sensor, where the amount of collected charge at the intercept is determined. Thedistribution of collected charge arising from repeating this process for many tracks is fittedwith a Landau distribution convolved with a Gaussian distribution to obtain the MPVof the Landau distribution. The EDV is then determined as the voltage at which theMPV of the charge distribution is equal to 80% of the MPV obtained at the maximumtest voltage (currently 200 V). This value of 80% provides good agreement between thedepletion voltage measured from this method prior to irradiation and that obtained fromthe capacitance-voltage measurements performed on the sensors at production. It alsorepresents a voltage above which the sensors must be operated to ensure significant signalcharge is extracted. This process then repeats via an automated procedure for a range ofpatterns such that all of the sensors are tested. As this procedure requires beam time thatwould otherwise be used for physics data taking, it is currently only performed aroundthree times per year.
Figure 24 shows the measured EDV values using all the VELO sensors, and dividingthem into radial regions with reasonably constant fluence. The delivered integratedluminosity at each CCE scan has been combined with the expected fluence per fb−1 to givethe fluence for each radial region of each sensor. The sensors exhibit an overall decreasein EDV with fluence followed by an increase in EDV until the most recent CCE scan(after a delivered integrated luminosity of 3.4 fb−1). As a result of the radiation damagethat had occurred by the start of 2013, many of the sensors now have EDVs significantlyhigher than before irradiation. On average, type inversion of the silicon occurs around20× 1012 neq, after which the EDV increases linearly with fluence. This behaviour cantherefore be used to predict the operational voltages required to give maximal CCE fora given fluence and hence a delivered integrated luminosity. It can also be noted that aminimum EDV of ∼ 20 V is observed for all sensors, this is assumed to be a result of thefinite integration time of the front-end electronics.
4.6.3 Charge loss to second metal layer
The cluster finding efficiency in the sensors is studied using the same technique of trackextrapolation as that used to study the CCE (see Sect. 4.6.2). These studies have shownan unexpected radiation induced charge loss due to the presence of the second metallayer on the sensors. The second metal layer on the sensors is used to route signals frominner strips to the outer radius of the sensor. Particles that pass close to a second metallayer trace, but away from the shielding effect of the first metal layer on the strip, havea reduced charge collected on this strip and additionally induce charge on the secondmetal layer. This gives rise to a reduction in cluster finding efficiency on this strip, andadditional noise clusters located on the strip connected to the second metal layer track.This degradation has been observed to continue throughout the 2012 data taking period,
38
fluenceeq1 MeV n0 20 40 60 80 100 120 140 160
1210×
Effe
ctiv
e de
plet
ion
volta
ge [V
]
0
20
40
60
80
100
120
LHCb VELO
2012 & 2013 data only
Figure 24: The effective depletion voltage versus fluence for all VELO sensors up to 3.4 fb−1
delivered integrated luminosity.
although there is some evidence to suggest the rate of change with fluence is decreasing.Currently studies of the VELO tracking efficiency show no degradation associated withthis effect within the errors of ±0.3%. This effect is discussed in detail in Ref. [30].
39
5 Physics performance
The performance of the VELO on more directly physics analysis related parameters isreviewed in this section. The performance of the pattern recognition, track reconstructionand the alignment of the detector are discussed. The impact parameter, primary vertex(PV) resolution and decay time resolution of the detector are then presented.
5.1 Pattern recognition and tracking
Obtaining a high efficiency for reconstructing the trajectories of charged particles isparticularly important for the analysis of many-body final states since the selectionefficiency scales with the track finding efficiency to the power of the number of tracks.An accurate knowledge of the tracking efficiency is also important in many other physicsanalyses, especially those that aim to measure a production cross-section or a branchingfraction. The track finding algorithms in LHCb start with a search in the VELO detectorfor straight lines. These tracks are then combined with hits in the tracking stationsto produce the “long” tracks that are used for most physics analyses. The track fit isperformed using a Kalman fit. In the trigger a simplified model of the detector materialand a single direction Kalman fit are used in order to reduce CPU consumption. Afull bi-directional Kalman fit is performed offline using a detailed model of the detectormaterial.
The VELO pattern recognition algorithm requires a minimum of three R sensor andthree Φ sensor clusters to reconstruct a trajectory. The basic algorithm collects sets ofR clusters consistent with being on a straight line from the interaction point, then looksfor a set of compatible Φ sensor clusters to confirm the trajectory. The alternating stereoangles of the Φ sensor strips resolves the stereo ambiguities when combining the R and Φsensor clusters. A second pass is made combining the unused clusters to find tracks notfrom the interaction region. The number of clusters on a VELO track ranges from six (theminimum requirement is three on R and three on Φ sensors) up to the full 42, with anaverage of 11. An additional algorithm, that only runs when the detector is closing, makes3D space points from the R and Φ sensor clusters in a module and forms tracks from thesehaving no assumptions about the track directions.
The efficiency of the track reconstruction in the VELO has been measured using atag-and-probe method. Samples of J/ψ decays into two muons are used where the track ofone of the muons is fully reconstructed (tag-muon), while the other muon is only partiallyreconstructed using hits in the other tracking stations (probe-muon). The momenta of thetag- and probe-muon are used to reconstruct the J/ψ candidate mass and hence confirmthe selection. It is required that the trigger has not selected the event based on the probemuon track to prevent any potential trigger bias. The VELO tracking efficiency is obtainedby matching the partially reconstructed probe muon to a long track (which has a VELOtrack segment).
The measured VELO tracking efficiency for long tracks is shown in Fig. 25 for data andsimulation, and is typically 98% or higher in the data. The simulation is weighted by the
40
)c (GeV/p0 50 100 150 200
ε
0.92
0.94
0.96
0.98
1
1.02
1.04Data 2011
MCVELO tracking efficiencyLHCb
η2 3 4 5
ε
0.92
0.94
0.96
0.98
1
1.02
1.04Data 2011
MCVELO tracking efficiencyLHCb
φ-2 0 2
ε
0.92
0.94
0.96
0.98
1
1.02
1.04Data 2011
MCVELO tracking efficiencyLHCb
trackN0 200 400
ε
0.92
0.94
0.96
0.98
1
1.02
1.04Data 2011
MCVELO tracking efficiencyLHCb
Figure 25: Tracking efficiency for the 2011 data and simulation for the VELO as a function ofthe momentum, p (top left), the pseudorapidity, η (top right), the azimuthal angle φ (bottomleft) and the total number of tracks in the event, Ntrack (bottom right). The simulation hasbeen reweighted to the number of tracks observed in data for the p,η and φ plots. The errorbars indicate the statistical uncertainty.
number of tracks observed in the data. The efficiency for the VELO pattern recognitionis weakly dependent on the multiplicity of the event, as the pattern recognition becomesmore complex in a higher occupancy environment. The dips in efficiency at |φ| ≈ π/2 arecaused by the extra multiple scattering from the material of the RF foil in the verticalplane, see Fig 5 (left). The discrepancy between data and simulation in these bins ispartially explained by the effect of assumptions made in the tracking: the beam is centredin the coordinate system in the simulation while it is offset by 0.4 mm in data; and thedistance between the two halves when the system is fully closed differ at the 150µm levelbetween data and simulation. The discrepancy is taken into account in physics analysesby applying a reweighting procedure that ensures data and simulation agreement, and aprocedure for determining the systematic uncertainties due to this correction is also inplace.
Another important measure of the tracking performance is the number of poor quality
41
Number of VELO clusters0 2000 4000
Fra
ctio
n o
f g
ho
sts
in A
ccep
tan
ce
0
0.1
0.2
LHCb VELO Simulation
Number of VELO Clusters0 2000 4000
Tim
e [m
s]
0
5
10
15
20
LHCb VELO Simulation
Figure 26: (left) Fraction of ghost tracks versus number of VELO clusters in simulation. (right)VELO pattern recognition timing versus number of clusters, the pattern recognition alone(magenta triangles), the raw data decoding (blue squares) and the combined (red circles) timingsare shown. The times are scaled to a 2.8 GHz Xeon processor.
or “ghost” tracks that are produced. Defining a ghost track as one in which less than70% of the VELO clusters on the track are from the same simulated particle, the fractionof ghost tracks is shown in Fig. 26 (left) as a function of the total number of VELOclusters. These tracks become more frequent as the detector occupancy rises, with typicaloccupancies being 0.5% for randomly triggered events and 1% for HLT triggered events(see Sect. 4.3).
The CPU resources consumed by the VELO tracking in the high level trigger aresubstantial and the algorithms have been optimised for speed. For example, to minimisethe time taken in decoding the clusters, the 3-bit cluster centre position computed inthe TELL1 FPGAs is used (see Sect. 3.3). The time taken for the decoding and patternrecognition algorithms is shown in Fig. 26 (right).
5.2 Alignment
The VELO has extremely stringent alignment requirements to ensure that the intrinsichit resolution of the detector (see Sect. 4.2) and the impact parameter and decay timeresolution (see below) of the experiment are not adversely affected. Furthermore, asdescribed in Sect. 2.5, the VELO halves are inserted and centred around the beams ineach LHC fill. Consequently the high level trigger requires an immediate update of thealignment parameters. These parameters are also required for the offline reconstruction.The alignment of the detector is thus separated into two elements: the underlying systemalignment, and the updates required in each fill.
The underlying alignment of the VELO relies on three components: the preciseconstruction and assembly of the detector, the mechanical and optical survey of each partof the detector, and the software alignment of the system using tracks. The updates for
42
each fill are then added to this: mechanical measurements of the position to which thedetector is closed are used to determine these.
5.2.1 Optical and mechanical measurements
Each component of the detector was surveyed at the various stages of the assembly using asmart-scope for the individual modules, and a coordinate measuring machine with opticaland touch probe heads for the modules and for each half of the assembled system. Therelative position of the Φ sensor with respect to the R sensor in each module was measuredwith an accuracy of about 3µm for the x and y translation and with an accuracy of about20µrad for rotations around the x and y axis. The relative module positions within eachhalf of the detector were measured with a precision of about 10µm for the translations alongx and y. Measurements of the mounting frame were made with a coordinate measuringmachine during production, and then prior to, and after, installation measurements weremade using photogrammetry, theodolites, mechanical gauges and levelling instruments.The position of the two VELO halves was determined with an accuracy of 100µm for thetranslations and 100µrad for the rotations. These survey measurements are used as astarting point for the track-based software alignment, and remain important for the finalalignment quality as some degrees of freedom (see below) are difficult to align with tracks.
5.2.2 Track-based alignment methods
The track-based alignment relies on minimising the residuals between the fitted tracksand the measured cluster positions. The alignment can be considered in terms of threedifferent stages:
1. the relative alignment of each Φ sensor with respect to the R sensor in the samemodule. This allows the x and y translations of the sensors to be determined;
2. the relative alignment of the modules within each VELO half. This alignment isprimarily sensitive to the x and y translations of the modules and their rotationsaround the z axis. Only the Φ sensors are sensitive to this rotation due to the stripgeometry and hence this rotation must be determined at the module level. Themisalignment due to the other three degrees of freedom (the z translation and therotations around the x and y axis) cause second order effects to which sensitivitycan only be obtained with R sensors from using a large data sample with a widerange of track angles;
3. the relative alignment of one VELO half with respect to the other half. The PVposition can be used as a constraint since tracks reconstructed in each half originatefrom this common point. This method is sensitive to the x, y and z translations andthe rotations around the x and y axis. In addition, tracks that cross both halves ofthe detector are used. There is a small overlap between the sensors in the left andright halves of the VELO when the detector is fully closed, and tracks that traversethe VELO in this overlap region are particularly important. The alignment using
43
these tracks is mainly sensitive to the misalignment of x and y translations and therotation around the z axis.
Two methods were developed and used for performing the track-based software align-ment. The first method [34,35] performs the Φ sensor alignment by fitting an analyticalform to the residuals as a function of φ, and performs the remaining alignment stages usinga matrix inversion method, based on Millepede [36], which performs a χ2 minimisationthat depends upon both the track and alignment parameters. This alignment method isfast but the implementation fits straight line tracks in the VELO. The second methoduses a global χ2 minimisation based on Kalman track fit residuals [25]. This method usesan iterative procedure, that could be time consuming in the case of a large misalignment.However, the Kalman filter fit takes into account the corrections for multiple scattering,energy loss effects and the weak magnetic field in the VELO region. The results from bothmethods are in agreement [14].
The alignment is performed using a specially selected data sample which improves thesensitivity to all degrees of freedom and better constrains the upstream and downstream ofthe PV regions of the VELO. This sample includes a mixture of tracks from collisions witha wide range of angles, and tracks from beam-gas interactions. The beam-gas interactionsare roughly parallel with the beam-axis, giving tracks that cross many modules.
Correlated module misalignments which are poorly constrained in the track-basedalignment, known as weak modes, can bias physics quantities like the impact parameteror the invariant mass. The most important weak mode in the VELO is the twist ofthe modules around the z-axis. This misalignment can distort the impact parametermeasurement by several tens of microns as a function of the azimuthal direction of thetrack. The effect is limited by using the alignment track sample with a range of track types,and by applying constraints in the alignment procedure from the survey measurements.
5.2.3 Mechanical measurement of closing
The VELO halves are moved independently horizontally (x) and together vertically (y)during the closing procedure and the motion is measured to an accuracy of a fewµm withmechanical position sensors. These measure the revolutions of the spindle that controlsthe displacement. The position of the two VELO halves is updated for each fill using thesemechanical system measurements in x and y by adding these changes to the underlyingsystem alignment.
Performing the track-based alignment of the VELO halves using data collected atdifferent opening positions (distance between the two halves of 0, ±5 mm, ±10 mm,±29 mm) allows a calibration of the motion system measurements to be performed. Thisshows a calibration accuracy of 0.6% of the motion system along the x direction is achieved.
5.2.4 Alignment performance
The track-based alignment results are in good agreement with the lower precision surveyresults, within the uncertainties, apart from the effect due to the change in temperature
44
Figure 27: Example unbiased sensor residuals as a function of the φ coordinate using only thesurvey information (left) and using the track-based software alignment (right). Results are givenfor two different example sensors. (top) A significant improvement in the residuals is seen in thissensor with the track-based alignment. (bottom) In this sensor the alignment quality using thesurvey information is already good.
of the system. The system was surveyed at room temperature, but is normally operatedwith the cooling system operated at -30◦C. Comparing track-based alignments performedwith the cooling temperature set to +8◦C and -30◦C a change of distance between thetwo halves of about 170µm was measured. Laboratory measurements were also madeby heating individual modules and an expansion of approximately 1µm per degree alongthe x direction of the carbon fibre that supports the modules was found. Hence, evenwhen using materials with a low coefficient of thermal expansion it is important to controltemperature changes when aiming for µm level precision. This is achieved in the VELOby maintaining the mounting base plate of each VELO half at 20◦C and operating with astable cooling system temperature.
The improvement obtained after the track-based alignment procedure over the precisionof the survey is illustrated in Fig. 27. The shapes of the residual distributions as a functionof the azimuthal coordinate are characteristic of particular sensor misalignments. Theprecision obtained after the track-based alignment is better than the best hit resolutionof the detector with an alignment at the few µm level obtained for the x and y moduletranslations.
Fitting the position of the PV separately with tracks in the two halves of the VELOallows the misalignment between the two halves to be determined. This is shown in Fig. 28over a period of four months of operation. The variation between runs shows the accuracywith which the position in each fill is measured. The excellent stability of the alignment ofthe system is also clear.
45
Run Number74000 76000 78000 80000
m]
µ) [
A-C
PVx
∆M
ean
(
-20
-15
-10
-5
0
5
10
15
20LHCb VELO
Figure 28: Misalignment between the two VELO halves in each run, evaluated by fitting thePV separately with tracks in the two halves of the VELO. The run numbers shown here spanthe period of the last four months of operations in 2010.
5.3 Primary vertex resolution
The accurate measurement of decay lifetimes is required for the primary physics aims ofthe LHCb experiment in CP violation and rare decay studies. Precise vertex reconstructionis therefore of fundamental importance, in order to resolve production and decay vertices.
The PV resolution is strongly correlated to the number of tracks N used to reconstructthe vertex. The analysis is performed on an event-by-event basis. The principle is toreconstruct the same PV twice, and to determine the difference between these two PVpositions. This is achieved by splitting the track sample of each event into two and makingvertices from each independent set of tracks. The method was verified in the simulationby comparing the reconstructed and generator level information.
The track splitting is done entirely at random, with no ordering of tracks and norequirement that the same number of tracks is put into each set. The vertex reconstructionalgorithm is applied to each set of tracks. Vertices are ‘matched’ between the two setsby requiring that the difference in their z position is < 2 mm. Then, if the number oftracks making a pair of matched vertices is the same, the residual is calculated. Repeatingfor many events yields a series of histograms of residuals in (x, y, z) for varying trackmultiplicity.
In practice, the number of tracks making a vertex ranges from 5 (the required minimum)to around 100. However, given the track splitting method roughly divides the total numberof tracks in two, it is difficult to measure the resolution past 40 tracks. Each residualhistogram is fitted with a Gaussian distribution. The resolution for each particular trackmultiplicity is calculated as the σ of the fitted Gaussian divided by
√2, as there are two
uncorrelated resolution contributions in each residual measurement.The resolution is fitted with a function which parametrises it in terms of N as follows:
σPV =A
NB+ C, (1)
46
N5 10 15 20 25 30 35 40
m]
µ
PV re
solu
tion
[
0
5
10
15
20
25
30
35
40
LHCb VELO
m)µ (A
Bm)µ (C
97.70.61-1.0
0.9± 0.01± 0.2±
x{m)µ (A
Bm)µ (C
92.70.59-1.5
0.8± 0.01± 0.2±
y{
N5 10 15 20 25 30 35 40
m]
µ
PV re
solu
tion
[
0
50
100
150
200
250
300
LHCb VELO
m)µ (A
Bm)µ (C
9230.69-16
7± 0.01± 1±
z {
Figure 29: PV Resolution of events with exactly one PV in 2011 data as a function of trackmultiplicity. (left) x (red) and y (blue) resolution and (right) z resolution. The fit parameters A,B and C for each coordinate are given.
where A,B,C are constants.In 2011 data it was found that a 25-track vertex has a resolution in the transverse plane
of 13µm, while the resolution in z is 71µm, as shown in Fig. 29. The 2011 simulation hada resolution approximately 2µm better than in the data. For data with an average numberof visible proton-proton interactions per bunch crossing of around 1.3, the average numberof tracks in a minimum bias event containing one PV is 55. The equivalent number inan event in which a candidate B decay has been reconstructed is 120. As the number ofreconstructed PVs in the event increases, the resolution degrades. The rate of degradationis approximately 5–10% per additional vertex. The vertex resolution results for 2012 dataare very similar.
The stability of the PV position has also been studied. In a single fill the PV positionwas found to move in x and y by not more than 4µm and 2µm respectively. Based onanalysing a three-month data sample from 2012, the PV position in the LHCb coordinatesystem was found to move by a maximum of 50µm (RMS=16µm). The VELO is centredaround the beam with a maximum variation of 20µm in x and 40µm in y. For aconservative estimate we apply a safety factor of two to the total variation, and determinethat the beam stability with respect to the VELO is better than 100µm.
5.4 Impact parameter resolution
The impact parameter (IP) of a track is defined as the distance between the track andthe PV at the track’s point of closest approach to the PV. The B and D mesons studiedin many LHCb analyses are long lived particles and hence their decay vertex is generallydisplaced from the PV. The tracks made by particles coming from the decay of long livedparticles therefore tend to have larger IPs than those made by particles produced at thePV. Consequently, cuts on the IP are very effective at excluding prompt backgrounds,and maximising the signal content of a data set. It is thus of great importance for an
47
experiment like LHCb to be able to measure IPs to a high precision, and to have a goodunderstanding of the effects contributing to the resolution of IP measurements.
The IP resolution is governed by three main factors: multiple scattering of particles bythe detector material; the resolution on the position of hits in the detector from whichtracks are reconstructed; and the distance it is required to extrapolate a track from its firsthit in the detector to the PV. An approximate analytical expression can be derived [12],
σ2IP =
(0.0136
pT
√x/X0[1 + 0.038 ln(x/X0)]
)2
r21 +∆2
02σ21 + ∆2
01σ22
(∆02 −∆01)2, (2)
where the first term is due to multiple scattering and the second term due to the detectorresolution. The particle has a transverse momentum pT and it passes through a piece ofmaterial, of thickness x, and with radiation length X0. The track is extrapolated over adistance ∆01 between the first hit on a track and the PV, and the distance from the PV tothe second hit is ∆02. The resolution of the first and second hits on the track are σ1 andσ2. This expression applies to a 1D IP measurement. Parametrising the resolution in theplane perpendicular to the beam, IPx and IPy, as a function of 1/pT one expects identical,roughly linear distributions with an x or y-intercept dependent on the detector resolution,and a gradient proportional to the material budget. In 3D geometry an IP has two degreesof freedom: three as it is a distance in 3D space, minus one from the requirement of beingcalculated at the point of closest approach to the PV. Due to the forward geometry ofLHCb the z component is negligible, and the IP measurement in 3D space is thus simplythe sum in quadrature of its x and y components. The mean offset of such a measurementfrom its true value is given by the resolution on the 1D components multiplied by
√π/2.
The vast majority of tracks reconstructed at LHCb are made by particles producedpromptly at the PV. The measured IP of such tracks is non-zero only due to the mea-surement resolution. Thus, the IP resolution can be measured by examining the widthof the IPx and IPy distributions for all tracks. To do this, only good quality long tracksfrom events with only one reconstructed PV are used. The PV is required to have at least25 tracks included in its fit to minimise the contribution of the vertex resolution to themeasured IP. Furthermore, the PV is refitted excluding each track in turn before its IP iscalculated so the track has no influence on the PV position. The IPx and IPy are thenplotted in bins of the variable of interest, such as 1/pT , and a fit of a Gaussian functionperformed in each bin. The width of the fitted Gaussian is taken as the resolution.
Figure 30 shows plots of the IP resolution versus momentum and 1/pT . The resolutionof IPx and IPy is almost identical. They are asymptotic at high pT , tending to ∼12µm,and depend roughly linearly on 1/pT . The performance of the VELO in this respect isexcellent, achieving IP resolutions of <35µm for particles with pT >1 GeV/c.
The distribution of material in the VELO is non-uniform, and this also affects IPresolutions. The excellent agreement with simulation shown in Fig. 30 was only obtainedafter careful study of the material distribution in the RF foil. In the region in whichthe two halves of the VELO overlap the two sides of the RF foil also overlap, greatlyincreasing the material density. This can be seen by measuring the IP resolution in bins
48
]-1c [GeV p0 10 20 30 40 50 60 70
m]
µIP
res
olut
ion
[
0
10
20
30
40
50
60
70
80
90
100
2012 data
xIP
yIP
LHCb VELO
]-1c [GeV T
p1/0 0.5 1 1.5 2 2.5 3
m]
µ r
esol
utio
n [
xIP
0
10
20
30
40
50
60
70
80
90
100
Tp = 11.6 + 23.4/σ2012 data,
Tp = 11.6 + 22.6/σSimulation,
LHCb VELO
Figure 30: IPx and IPy resolution as a function of momentum (left) and IPx as a function of1/pT and compared with simulation (right). Determined with 2012 data.
[rad.]φ-3 -2 -1 0 1 2 3
m]
µ r
esol
utio
n [
xIP
0
10
20
30
40
50
60
70
80
90
100
2011 dataSimulationLHCb VELO
Figure 31: IPx resolution as a function of azimuthal angle φ, measured on 2011 data andcompared to simulation.
of the azimuthal angle φ, as is shown in Fig. 31. The increase in material is reflected inthe increase in IP resolution about φ = ±π/2, i.e. in the overlap region.
Thus, it can be seen that the VELO provides accurate IP measurements on which theLHCb physics programme relies for the rejection of prompt backgrounds to long-livedheavy flavour hadron decays. The IP resolution behaves as expected, with a roughlylinear dependence on 1/pT , and a clear dependence on both the hit resolution and thedistribution of material.
49
5.5 Decay time resolution
The reconstructed decay time of strange, charm and beauty hadrons is used in offline eventselections and in precise measurements of lifetimes. However, the most stringent demandson the decay time resolution stem from the requirement to resolve the fast oscillationsinduced by B0
s–B0s mixing. Consequently, we illustrate the performance of the VELO with
an analysis of the decay time resolution of B0s→ J/ψφ decays.
The reconstructed decay time in the rest frame of the decaying particle can be expressedin terms of the reconstructed decay length l, momentum p and mass m of the particle inthe LHCb frame as
t =m l
p. (3)
The decay time is computed with a vertex fit that constrains the decaying particle tooriginate from the PV. The decay time uncertainty is a function of the actual decay time.For small decay times it is dominated by the resolution on the decay length l, which inturn is dominated by the secondary vertex resolution of the VELO. For large decay times,it is dominated by the momentum, which is mostly determined by the tracking stationsbefore and after the LHCb magnet. The two contributions are approximately equal atseveral times the B hadron lifetime. Therefore, the momentum resolution plays only asmall role.
Time dependent CP violation effects are measured as the amplitude of an oscillationin the B decay time distribution. The size of the observed amplitude is damped by adilution factor from the finite decay time resolution. For a resolution function that is asingle Gaussian function with RMS σt, the dilution is
D = exp
[−1
2∆m2σ2
t
], (4)
where ∆m is the mixing frequency.The sensitivity to the oscillation amplitude is proportional to the dilution. Therefore,
for optimal sensitivity the dilution must be as close to unity as possible. However, evenmore important is the understanding of the dilution itself, since a bias in the estimateddilution leads to a bias in the measurement of a CP violating effect.
The decay time resolution depends on the topology of the decay and is calibrated foreach final state on data. The calibration method uses prompt combinations that fakesignal candidates. Ignoring the small contribution from signal candidates and long-livedbackground, the shape of the prompt peak is determined only by the resolution function.Figure 32 (left) shows the decay time distribution for fake B0
s → J/ψφ → µ+µ−K+K−
decays in data and simulation. The B0s → J/ψφ candidate selection is described in
Ref. [8]. The contribution from signal decays (which would result in a tail on the right ofthe distribution) is removed. The RMS values quoted in the figure are computed usingevents with negative decay time or decay length only, to reduce the sensitivity to a smallcontribution from other B → J/ψX decays. For a mixing frequency of 17.7 ps−1, the decaytime resolution corresponds to a dilution of about 0.7.
50
decay time [ps]-0.2 0 0.2
even
ts /
0.01
ps
0
10
20
310×data, rms = 50 fs
mc, rms = 44 fs LHCb
decay length [mm]-1 -0.5 0 0.5 1
even
ts /
0.05
mm
0
20
40
310×
mµmc, rms = 210 mµdata, rms = 230
LHCb
Figure 32: Decay time (left) and decay length (right) distribution for fake, prompt B0s → J/ψφ→
µ+µ−K+K− candidates in 2011 data (black points) and simulation (solid red histogram). Onlyevents with a single PV are used. The simulated data are generated inclusive J/ψ eventsfrom which signal B0
s → J/ψφ are removed. In the data contributions from non-J/ψ di-muoncombinations and from true B0
s → J/ψφ are subtracted using the sPlot technique [37].
The resolution in both decay time and decay length is about 10% worse in 2011 datathan in the corresponding simulated events. Note that the decay time resolution is moreGaussian shaped than the decay length resolution. This is largely because, for a fixedopening angle resolution, the decay length resolution is proportional to the momentum.This momentum dependence largely cancels in the decay time.
This effect is illustrated in Fig. 33 (left) which shows the resolution as a function of the(fake) B candidate momentum, where a mixing frequency of 17.7 ps−1 and Eqn. 4 havebeen used to obtain the decay time resolution. Only events with t < 0 were used in orderto reduce the sensitivity from remaining long-lived candidates. This procedure gives aresolution that is most relevant for the time-dependent CP violation measurement in B0
s
decays.Finally, Fig. 33 (right) shows the resolution as a function of the per-event estimated
uncertainty in the decay time. The latter is computed with the same vertex fit thatcomputes the decay time itself. It is a non-trivial function of the track parametersand covariance matrices of the four final state particles in the decay. As expected, theresolution is a linear function of the estimated uncertainty. However, since track parameteruncertainties are not perfectly calibrated yet in the data, the slope is ∼ 1.2 rather thanunity. The typical decay time resolution in LHCb is 50 fs, and this resolution plays acrucial role in the sensitivity of many LHCb physics measurements.
51
]c [GeV/p0 100 200 300 400 500
reso
lutio
n [f
s]
0
10
20
30
40
50
60
70
80
90
100
2011
2012
LHCb
[fs]tσ20 40 60 80
reso
lutio
n [f
s]
0
10
20
30
40
50
60
70
80
90
100
2011
2012
LHCb
Figure 33: Decay time resolution (points) as a function of momentum (left) and as a functionof the estimated decay time uncertainty (right) of fake, prompt B0
s → J/ψφ → µ+µ−K+K−
candidates in data. Only events with a single PV are used. Contributions from non-J/ψ di-muon combinations and from true B0
s → J/ψφ are subtracted using the sPlot technique. Thesuperimposed histogram shows the distribution of momentum (left) and estimated decay timeuncertainty (right) on an arbitrary scale.
52
6 Conclusions
The performance of the LHCb VELO during its first years of operation has been described.The operational experience of key subsystems has been reviewed. The sensors have beensuccessfully operated in a secondary vacuum at a pressure of 2× 10−7 mbar, and cooledusing a bi-phase CO2 cooling system which has maintained the operating temperature ofthe sensors at (−7± 2)◦C. The sensors are moved in 210 seconds from their fully retractedposition to be centred 7 mm from the LHC beam for physics operation to an accuracy ofbetter than 4µm. The average material budget of the detector for tracks in the LHCbacceptance is 0.22 X0.
The determination of the sampling time of the front-end pulse-shape and the digitisationtime to maximise the signal, minimise spillover to the next and preceding event, andsynchronise with the LHC beam collisions is described. The timing is found to be stableover a period of one year of operation at the nanosecond level. The normalisation of thegain of the system has also been described and this calibration is performed approximatelyevery six months. The digitised data are processed by a series of FPGA algorithms thathave been described and perform pedestal subtraction, common mode suppression andclusterisation. Errors in the system are also identified in the FPGA algorithms. A noveltechnique is applied of emulating the algorithms in bit-perfect C code which is usedin the main analysis framework of the experiment to tune the operational parameters.A parameter retuning is required approximately every two months to ensure optimalperformance. Single event upsets are observed in the front-end electronics at a rateof around 2.9 register bit-flips for all front-end ASICs combined per pb−1 of deliveredintegrated luminosity. Extensive data quality monitoring has been put in place, includingautomatic processing of each run, the use of a graphical user interface to display plots,and templated reports in an electronic logbook.
The sensors initially had a signal to noise ratio of approximately 20:1, with the noisedepending on the strip capacitance. The hit resolution varies with pitch and track angle;for the optimal track angle of 7–11◦ it varies from 4µm in the 40µm pitch region to20µm at 100µm pitch. This resolution has been achieved using analogue readout for thefront-end ASICs and pulse-height weighted cluster position determination. The typicalcluster occupancy in the experiment during 2011 operation was around 0.48% for randomlytriggered events in beam-crossings, but rises to 0.93% for events that have passed the highlevel trigger. Faulty strips in the detector have been determined using noise distributions,occupancy spectra and the cluster finding efficiency. The detector has less than 1% of faultystrips. Tracks arising from interactions between the LHC beams and gas molecules areobserved and provide a useful sample for alignment and for beam imaging for luminositystudies. Beam backgrounds giving rise to splashes of hits in localised regions of the detectorare also seen. The radiation damage in the detector has been studied through the analysisof the currents drawn and charge collection versus voltage. Dedicated beam time is usedfor the latter study for which an automated procedure has been developed. The innerradius regions of the n+-on-n sensors are observed to have undergone space-charge signinversion, which is expected due to their proximity to the LHC beams. Charge loss is also
53
observed to have developed after irradiation due to the presence of the second metal layerthat is used for routing out the signals from strips on the sensor.
The track finding efficiency of the VELO has been determined using a tag and probemethod using J/ψ decays, and is typically above 98%. The modules have been alignedusing a track-based software alignment procedure. The position to which the VELO isinserted in each fill is measured by mechanical motion sensors and this is used to updatethe alignment. An alignment precision of 1µm for translations in the plane transverse tothe beam, and a 5µm stability of the relative alignment of the two VELO halves over anoperational year is obtained.
The LHCb physics selection and analysis of long-lived heavy flavour decays relies onthe background rejection and flavour tagging from the impact parameters and vertexingperformance of the detector. The vertex resolution is strongly dependent on the numberof tracks in the vertex. A resolution of 13µm in the transverse plane and 71µm along thebeam axis is achieved for vertices with 25 tracks. A 1D impact parameter resolution of12µm in the plane transverse to the beam for high momentum tracks is obtained. It isprimarily determined by the detector cluster position resolution and the distance of thesensors for the LHC beams. For lower momentum tracks the impact of multiple scatteringin the detector material becomes dominant, and an impact parameter resolution of 35µmis achieved for particles with transverse momentum of 1 GeV/c. A decay time resolutionof approximately 50 fs is obtained (evaluated for the B0
s→ J/ψφ decay channel) whichplays a key role in many LHCb physics results.
7 Acknowledgements
This complex detector could only be constructed with the dedicated effort of manytechnical collaborators in the institutes forming the LHCb VELO Collaboration. A specialacknowledgement goes to all our LHCb collaborators who over the years have contributedto obtain the results presented in this paper. We express our gratitude to our colleaguesin the CERN accelerator departments for the excellent performance of the LHC. Wethank the technical and administrative staff at the LHCb institutes. We acknowledgesupport from CERN and from the national agencies: CAPES, CNPq, FAPERJ andFINEP (Brazil); NSFC (China); CNRS/IN2P3 and Region Auvergne (France); BMBF,DFG, HGF and MPG (Germany); SFI (Ireland); INFN (Italy); FOM and NWO (TheNetherlands); SCSR (Poland); MEN/IFA (Romania); MinES, Rosatom, RFBR and NRC“Kurchatov Institute” (Russia); MinECo, XuntaGal and GENCAT (Spain); SNSF and SER(Switzerland); NAS Ukraine (Ukraine); STFC (United Kingdom); NSF (USA). We alsoacknowledge the support received from the ERC under FP7. The Tier1 computing centresare supported by IN2P3 (France), KIT and BMBF (Germany), INFN (Italy), NWO andSURF (The Netherlands), PIC (Spain), GridPP (United Kingdom). We are indebted tothe communities behind the multiple open source software packages we depend on. We arealso thankful for the computing resources and the access to software R&D tools providedby Yandex LLC (Russia).
54
References
[1] LHCb Collaboration, A. A. Alves Jr. et al., The LHCb detector at the LHC, JINST 3(2008) S08005.
[3] LHCb Collaboration, R. Aaij et al., Measurement of σ(pp→ bbX) at√s = 7 TeV in
the forward region, Phys. Lett. B694 (2010) 209, arXiv:1009.2731.
[4] T. Sjostrand, S. Mrenna, and P. Z. Skands, PYTHIA 6.4 physics and manual, JHEP0605 (2006) 026, arXiv:hep-ph/0603175.
[5] R. Aaij et al., The LHCb trigger and its performance in 2011, JINST 8 (2013) P04022,arXiv:1211.3055.
[6] LHCb Collaboration, R. Aaij et al., Measurements of the branching fractionsfor B(s) → D(s)πππ and Λ0
b → Λ+c πππ, Phys. Rev. D84 (2011) 092001,
arXiv:1109.6831.
[7] LHCb Collaboration, R. Aaij et al., Precision measurement of the B0s − B0
s os-cillation frequency with the decay B0
s → D−s π+, New J. Phys. 15 (2013) 053021,
arXiv:1304.4741.
[8] LHCb Collaboration, R. Aaij et al., Measurement of the CP -violating phase φs in thedecay B0
s → J/ψφ, Phys. Rev. Lett. 108 (2012) 101803, arXiv:1112.3183.
[9] M. Agari et al., Beetle – a radiation hard readout chip for the LHCb experiment, Nucl.Instrum. Meth. A518 (2004) 468.
[10] G. Haefeli et al., The LHCb DAQ interface board TELL1, Nucl. Instrum. Meth. A560(2006) 494.
[11] A. Bates et al., LHCb vertex locator module characterisation and long term qualityassurance tests, Nucl. Instrum. Meth. A611 (2009) 41.
[12] A. Papadelis, Characterisation and commissioning of the LHCb VELO detector, PhDthesis, Vrije Universiteit, Amsterdam, June, 2009, CERN-THESIS-2009-044.
[13] C. Parkes, S. Borghi et al., First LHC beam induced tracks reconstructed in the LHCbVELO, Nucl. Instrum. Meth. A604 (2009) 1.
[14] S. Borghi, LHCb vertex locator commissioning and results from the first LHC beamreconstructed tracks, Nucl. Instrum. Meth. A623 (2010) 156.
[15] M. Clemencic, N. Gilardi, and J. Palacios, LHCb conditions database, LHCb-2006-017.
[20] M. Ferro-Luzzi, T. Latham, and C. Wallace, Determination of the aperture of theLHCb VELO RF foil, LHCb-PUB-2014-012.
[21] A. Bay et al., The LHCb VELO analogue transmission line, Nucl. Instrum. Meth.A614 (2010) 34.
[22] T. Szumlak and C. Parkes, Description of the Vetra Project and its Application forthe VELO Detector, LHCb-2008-022. CERN-LHCb-2008-022.
[23] H. Bichsel, Straggling in thin silicon detectors, Rev. Mod. Phys. 60 (1988) 663.
[24] R. Fruhwirth, Application of Kalman filtering to track and vertex fitting, Nucl. Instrum.Meth. A262 (1987) 444.
[25] W. Hulsbergen, The global covariance matrix of tracks fitted with a Kalman filterand an application in detector alignment, Nucl. Instrum. Meth. A600 (2009) 471,arXiv:0810.2241.
[26] S. D. B. Goddard, Proceedings of the 2012 Evian Workshop on LHC Beam Operation,CERN-ATS-2013-045.
[27] R. B. Appleby et al., Simulation of machine induced background in the LHCb experi-ment: methodology and implementation, IEEE Trans. Nucl. Sci. 59 (2012) 1681.
[28] LHCb Collaboration, R. Aaij et al., Absolute luminosity measurements with the LHCbdetector at the LHC, JINST 7 (2012) P01010, arXiv:1110.2866.
[29] C. Barschel, Precision luminosity measurement at LHCb with beam-gas imaging, PhDthesis, RWTH Aachen University, March, 2014, CERN-THESIS-2013-301.
[30] A. Affolder et al., Radiation damage in the LHCb vertex locator, JINST 8 (2013)P08002, arXiv:1302.5259.
[31] A. Gureja et al., Use of IV (current vs voltage) scans to track radiation damage inthe LHCb VELO, LHCb-PUB-2011-020.
[32] A. Hickling et al., Use of IT (current vs temperature) scans to study radiation damagein the LHCb VELO, LHCb-PUB-2011-021.
[33] M. Moll, Radiation damage in silicon particle detectors, PhD thesis, University ofHamburg, 1999, DESY-THESIS-1999-040.
[34] S. Viret, C. Parkes, and M. Gersabeck, Alignment procedure of the LHCb vertexdetector, Nucl. Instrum. Meth. A596 (2008) 157, arXiv:0807.5067.
[35] M. Gersabeck, S. Viret, C. Parkes et al., Performance of the LHCb vertex detectoralignment algorithm determined with beam test data, Nucl. Instrum. Meth. A596(2008) 164, arXiv:0807.5069.
[36] V. Blobel and C. Kleinwort, A new method for the high precision alignment of trackdetectors, arXiv:hep-ex/0208021.
[37] M. Pivk and F. R. Le Diberder, sPlot: a statistical tool to unfold data distributions,Nucl. Instrum. Meth. A555 (2005) 356, arXiv:physics/0402083.
![Plans and Status of the LHCb Upgrade - Indico · 2018. 11. 21. · Vertex Locator (VELO) IV [LHCB-TDR-013] Readout front-end chip –VeloPix ASIC Each sensor (43×15 I I) bump-bonded](https://static.documents.pub/doc/80x56/60fa5d20f20891506e1dde6c/plans-and-status-of-the-lhcb-upgrade-indico-2018-11-21-vertex-locator-velo.jpg)
